@comchatx/cx-sponsored-ads-playbook 0.0.0-stage → 0.1.0

package/research/beyond/brand-defense-and-competition.md

@@ -0,0 +1,872 @@
+# Brand defence and competition on the results page
+
+Read 2026-11-09 by the research seat. Tags: [official] Amazon's own words (Amazon Ads help pages, whose body is served inside the page's own HTML in `window.__scV2Content` and was decoded from it; Amazon Ads API guides read through their markdown sources at `https://d3a0d0y2hgofx6.cloudfront.net/en-us/<path>.md`; the merged OpenAPI model on Amazon's CloudFront host; release notes; Amazon Ads library, case-study and "what's new" pages; sell.amazon.com; aboutamazon.com; SP-API models on `amzn/selling-partner-api-models` and the SP-API report-type page), plus the US Federal Trade Commission's press releases on its sponsored-ads case, tagged [official, regulator — allegations, untested]; [opinion] a named practitioner, agency or paid tool, with who says it; [research] published research, with what was measured and on what data; [legal] a court decision, read through a public summary where the opinion itself was not reachable; [seat] the research seat's own reasoning, allowed here, never a position.
+
+Figures marked **(human readers only)** are dated snapshots. A skill does not carry them as rules; it reads the current value from the page or the API at run time.
+
+## What this covers
+
+**The part of a seller's ad work.** Everything a seller's ads do about other brands on the same results page and the same detail page, and everything other brands' ads do about the seller's. In order:
+
+- **Bidding on one's own brand terms** — what Amazon calls the *Protect* strategy and "brand defense": why a seller pays for a query whose shopper already typed the seller's name, what Amazon says about it, and what the controlled experiments outside Amazon measured about how much of that traffic is incremental.
+- **Defending one's own detail pages** — targeting one's own ASINs with Sponsored Products, Sponsored Brands and Sponsored Display so that the paid slots on the seller's own page carry the seller's catalogue, and the negatives that keep a seller's ads off pages it does not want.
+- **Conquesting** — targeting competitors' brand names as keywords and competitors' ASINs as product targets; and the policy line Amazon draws between a keyword (a targeting choice the shopper never sees) and creative (what the shopper reads).
+- **Share metrics** — search term impression share and rank, top-of-search impression share, the console's Brand Impression Share dashboard, the share-of-voice insight for partners, and the brand-share columns of Brand Analytics' Search Query Performance.
+- **Brand Analytics as a competitive signal** — Top Search Terms (the three most-clicked ASINs per query) and Search Query Performance (the brand's share of a query's impressions, clicks, cart adds and purchases).
+- **Sponsored Brands and the Store as brand tools** — the "Grow brand impression share" goal, Reserve share of voice (a fixed-price reservation of the Sponsored Brands top-of-search slot on branded keywords), theme targeting on "keywords related to your brand", Store Spotlight, branded-search metrics and peer benchmarks.
+- **Competitive response and price wars in the auction** — how Amazon describes its auction and reserve prices, the 2026 FTC case about them, and what the research on bidding cycles and competitive poaching says about escalation.
+- **Stock-outs and the Featured Offer as competitive moments** (gap fill) — what Amazon says happens to a product's ads when it goes out of stock or its SKU loses the Featured Offer; what that means for a rival's ads and for one's own when a reseller holds one's Featured Offer; the Keepa and SP-API reads of a rival's stock and Featured Offer state; and the practitioners who change bids on those events (§K).
+- **Sponsored Brands video and Store Spotlight in defence and conquest** (gap fill) — where Amazon says each format serves, how each is targeted, the creative rules that bound a video shown on a rival's term or page, and what practitioners report about video on competitor terms (§L).
+- **The research on competition in sponsored search** — brand-keyword experiments (Blake, Nosko & Tadelis; Simonov, Nosko & Rao; Simonov & Hill; Coviello, Gneezy & Goette), competitor-side spillovers (Golden & Horton), game-theory models of buying a rival's keyword (Desai, Shin & Staelin; Sayedi, Jerath & Srinivasan), and trademark-policy natural experiments (Bechtold & Tucker; Chiou & Tucker).
+
+**Why a senior Amazon advertising lead cares.** Three facts shape this whole area. First, Amazon allows any advertiser to target any brand's name as a keyword and any ASIN as a product target, so a seller's brand query and a seller's detail page are open auctions in which competitors buy space (§A, §F). Second, Amazon never names who those competitors are: every share metric it publishes is the seller's own share, or a rank, against an unnamed field (§D); competitor identity comes only from observing the page or from Brand Analytics' top-clicked products (§G). Third, the controlled evidence says that paying for one's own brand query buys few extra clicks when nobody else is bidding, and protects a measurable share when others are (§Research); so the value of brand defence depends on the competitive state of that query, which moves.
+
+**What this file does not cover, and where it lives.** Match types, the targeting expressions themselves and keyword research are in `areas/keyword-and-product-targeting.md` (whose §P on policy and §K on reserve share of voice this file builds on). Bid mechanics, placement multipliers and top-of-search bidding are in `areas/bids-and-placements.md` (§J there on top-of-search impression share). Campaign layout, including whether branded and non-branded terms share a campaign, is in `areas/campaign-structure.md`. Negative keywords as a cleaning tool are in `areas/search-terms-and-negatives.md`. Budgets in a bidding war are in `areas/budgets-and-pacing.md`. Prime Day and deal events are in `areas/seasons-and-deals.md`. Whether ad sales replace organic sales is in `areas/ads-organic-and-total.md`. Reading a sudden CPC jump or share loss is in `areas/diagnosis.md`. Which ad products a seller can buy is in `beyond/ad-products-and-eligibility.md`; metric definitions (new-to-brand, branded searches, halo) in `beyond/metrics-and-attribution.md`; what a seller is trying to earn from defence spend in `beyond/goals-and-economics.md`; Amazon's own automated tools in `beyond/automation-and-amazon-tools.md`. The API calls are in `method/ads-api-surface.md`, report mechanics in `method/reporting.md`, and how to read the effect of a change (including a brand-campaign pause) in `method/change-measurement.md`.
+
+## Amazon's own rules and guidance
+
+Each item is marked **RULE** (a constraint Amazon enforces: rejection, removal, a field that cannot change) or **GUIDANCE** (Amazon's advice, which a seller may ignore). Quotes are verbatim where the exact words carry weight.
+
+### A. Branded keywords: what Amazon says about bidding on one's own brand
+
+1. **GUIDANCE — Amazon names brand defence as a strategy.** *A guide to targeting with Sponsored Products* (advertising.amazon.com/library/guides/targeting-with-sponsored-products, read 2026-11-09) lists six targeting strategies: "Expand", "Promote", "**Protect** : focus on branded keywords and targeting", "**Conquest** : target competitors' products", "Upsell" and "Cross-sell". On branded keywords it says: "Branded keywords are directly associated with your brand or a brand you sell… Consider bidding on your branded keyword as a brand defense strategy since your competitors may be bidding on your branded keywords, too." [official]
+
+2. **GUIDANCE — what counts as a branded keyword.** *What is Keyword Targeting? How It Works, Benefits, Examples* (advertising.amazon.com/library/guides/keyword-targeting, read 2026-11-09): "Branded keywords include brand names, product lines, or trademark terms… Consider including variations of your brand and product names." [official] For reserve share of voice the help page defines the class more tightly (§B.2): misspellings and variations are rejected there.
+
+3. **GUIDANCE — Amazon's own figures on branded search (human readers only).** The unBoxed 2025 announcement (advertising.amazon.com/library/news/unboxed-2025-sponsored-brands-reserve-share-of-voice, dated November 11, 2025, byline Kelly Lui, read 2026-11-09): "70% of Amazon shoppers begin their journey with search, and 1 in 5 searches include a brand name", and "advertisers only achieve about 62% Sponsored Brands top-of-search impression share for branded queries on average." [official] *Your complete guide to Sponsored Brands* (advertising.amazon.com/library/guides/sponsored-brands-what-to-know, read 2026-11-09): "More than a quarter of shoppers shop across multiple brands, so it's important to be competitive against both category-level shopping queries and branded shopping queries." [official]
+
+4. **GUIDANCE — Sponsored Brands for brand up-selling.** The same Sponsored Brands guide: "Use individual product targeting if you are interested in brand up-selling and cross-selling." [official]
+
+5. **GUIDANCE — "keywords related to your brand" as a theme.** In Sponsored Brands theme targeting, `KEYWORDS_RELATED_TO_YOUR_BRAND` is "keywords that are often used to search for brands and promote brand impressions", beside `KEYWORDS_RELATED_TO_YOUR_LANDING_PAGES` (guides/sponsored-brands/targeting/theme-based-targeting.md, read 2026-11-09) [official]. The merged v1 OpenAPI model's `ThemeMatchType` enum (d1y2lf8k3vrkfu.cloudfront.net/openapi/en-us/dest/AmazonAdsAPIALLMerged_prod_3p.json, read 2026-11-09) describes it as "Search terms related to your brand", and carries a second brand-facing theme: `KEYWORDS_RELATED_TO_PEER_BRANDS_PRODUCT_CATEGORY` — "Search terms that shoppers often use when searching for and interacting with products from other brands in the category of your advertised products. **The peer brands are selected automatically.**" [official] `ThemeMatchType` is one enum shared across ad products (it also carries `KEYWORDS_CLOSE_MATCH`, `INTERESTED_AUDIENCE` and the like), and neither the model nor the two theme-targeting guides read (SB theme-based targeting; SP theme-based bid suggestions) say which ad product accepts the peer-brands theme; the SB theme guide names only `KEYWORDS_RELATED_TO_YOUR_BRAND` and `KEYWORDS_RELATED_TO_YOUR_LANDING_PAGES`. [seat] So Amazon defines an automated conquesting theme whose competitor list it chooses and does not show; where it can be used is open (Open question 15).
+
+6. **GUIDANCE — Amazon labels a recommended keyword as branded or not.** The recommendations API's `BRAND_PRODUCT_KEYWORD` type (Sponsored Brands, `POST /adsApi/v1/query/recommendations`, scoped by `brandId`/`brandName`) returns a `keywordTheme` per keyword, with values such as `Generic`, `Branded` and `Seasonal` (guides/recommendations/recommendations-api/recommendation-types.md, read 2026-11-09). [official] This is the only Amazon-side branded/non-branded label on a keyword that the seat found in the API; reports carry no such flag on a search term.
+
+7. **RULE — branded-search metrics match brand names, trademarks and variations, and do not exist for Sponsored Products.** *Branded searches metric enhancements* (advertising.amazon.com/resources/whats-new/branded-searches-metric-enhancements, read 2026-11-09): from October 1, 2025, "Enhanced brand matching" counts searches containing a "brand name, trademark, or common variations like abbreviations and misspellings"; "These metrics are not available for Sponsored Products"; they are available for Sponsored Brands, Sponsored Display, Sponsored TV and DSP. [official] The v1 field `metric.brandedSearches` is "The number of times shoppers searched for a brand promoted in the campaign, attributed to an ad interaction", eligible for SB, SD, STV, DSP and Full-Funnel (guides/reporting/ads-v1/metrics/amazon-retail-conversions/branded-searches.md, read 2026-11-09); its family includes `metric.costPerBrandedSearch`, `metric.brandedSearchRate` (branded searches divided by impressions) and `metric.brandedSearchesFromClicks` (only the P25/P50/P75 benchmark variants such as `metric.costPerBrandedSearchP25` list Sponsored Products among their programs). The v3 column `brandedSearches` is described differently: "The number of searches that included the name of your brand occurring within 14 days of an ad click or view" (guides/reporting/v3/columns.md, read 2026-11-09). The 2026-08-26 release note adds new-to-brand branded searches and Brand Store page views to v1 (release-notes/index.md, read 2026-11-09). [official] Detail in `beyond/metrics-and-attribution.md`.
+
+### B. Reserve share of voice: renting the Sponsored Brands top slot on one's own brand
+
+Reserve share of voice (RSOV) is a Sponsored Brands campaign goal that buys a fixed share of the Sponsored Brands top-of-search placement on the seller's own branded keywords, at a price fixed before the flight. It is the only Amazon product that sells a guaranteed share of a results-page slot to self-service advertisers. `areas/keyword-and-product-targeting.md` §K covers its targeting side; this section carries the rules that make it a brand-defence tool.
+
+1. **RULE — three Sponsored Brands goals, one per campaign, fixed.** *Campaign goals for Sponsored Brands* (help GXXPFXLSVX9VD3U9, lastModified 2026-08-11, read 2026-11-09) lists "Drive page visits" (CPC, all placements: search pages, detail pages, Amazon homepage), "Grow brand impression share" (vCPM, top-of-search) and "Reserve share of voice" (fixed price, top-of-search); one goal per campaign, and it cannot be changed after launch. The SB v4 OpenAPI (`openapi/en-us/sponsored-brands/4-0/openapi.json`, read 2026-11-09) names the API values: `BRAND_IMPRESSION_SHARE` ("This goal will allow you grown [sic] your brand impression share on top of search placement") and `PAGE_VISIT` (the default), with `ACQUIRE_NEW_CUSTOMERS` and `AD_VIEWS` marked "PREVIEW ONLY", and "Goal-based campaigns are not supported in the Ireland (IE) marketplace." (in IE, supplying `goal` and `costType` returns `GOAL_BASED_CAMPAIGNS_DISABLED`). [official]
+
+2. **RULE — only one's own registered brand terms can be reserved.** *Branded keyword guidelines and keyword suspension* (help G2QZJUGUT4RGLJ6N, lastModified 2025-11-05, read 2026-11-09): reserved keywords must be branded ("Glow moisturizer" is accepted; "moisturizer" is not). Rejection reasons, verbatim: the brand or trademark is not in Brand Registry; the keyword is too broad; "common misspellings and variations"; non-branded keywords; "A keyword contains both your and another company's brand name or trademark or your keyword has a brand name or trademark shared by more than 2 brands"; "You used keywords that contain another company's brand name or trademark"; the brand is not mapped to the account. [official] The library guide *How to set up Amazon Sponsored Brands reserve share of voice* (advertising.amazon.com/library/guides/sponsored-brands-reserve-share-of-voice, read 2026-11-09) answers the direct question: "Can I reserve generic or competitor keywords? No. Only exact match branded keywords… Generic terms, misspellings, competitor brands, and over-broad shared trademarks are excluded." [official] So the guaranteed slot can be bought only for defence, never for conquest.
+
+3. **RULE — keyword hygiene in RSOV.** Same help page: "You cannot reserve the same keyword on multiple campaigns or ad groups" (moving one means archiving it in the old ad group first); keywords must be mapped to a relevant ad within 14 days or they are removed, with charges for up to 14 days; "Once your keyword is removed, you cannot add it back to the same campaign"; if the account is suspended for non-payment, the ads stop. [official] Figures human readers only.
+
+4. **RULE — what is frozen at launch.** *Reserve keywords in a Sponsored Brands campaign* (help G86SD7HK6NHHRB9B, lastModified 2026-07-07, read 2026-11-09): "Once the campaign starts, you can't change your keywords, campaign duration, or budget." Suggested keywords are "listed in order of keyword search volume". Moderation usually finishes within 24 hours and can take up to 3 business days. "Note: Reserve share of voice is not available in Spain." (the same note is on the goals page GXXPFXLSVX9VD3U9, re-read 2026-11-09). [official]
+
+5. **RULE, figures conflicting across Amazon's pages (human readers only).** Flight length: 15–92 days, start 3 days to 3 months ahead (help G86SD7HK6NHHRB9B); "between 30 to 92 days" (goals page GXXPFXLSVX9VD3U9); "e.g., 14–90 days" (library guide). Minimum: "A minimum budget of $3,000 is required" (help G86SD7HK6NHHRB9B); "minimum spend of $6K per campaign for the U.S. marketplace" (library guide FAQ). Keyword count: 5–1,000, the one figure on which all three help pages agree (G2QZJUGUT4RGLJ6N: "You must reserve a minimum of 5 keywords and a maximum of 1,000"; G86SD7HK6NHHRB9B: "Reserve 5 to 1,000 keywords"; GXXPFXLSVX9VD3U9: "5 to 1,000 branded keywords for your own brands"; all re-read 2026-11-09); the v1 pricing schema accepts 1–1,000 keywords per pricing request (`SBBrandedKeywordsPricingCreate.keywords`, minItems 1, maxItems 1000), and the API guide requires at least five deal targets. Marketplaces: the library guide lists 20 (US, CA, MX, BR, DE, FR, IT, UK, BE, PL, NL, SE, TR, IE, AU, JP, IN, SG, AE, SA) while the SB v4 model says goal-based campaigns are not supported in IE and two help pages (G86SD7HK6NHHRB9B and GXXPFXLSVX9VD3U9) say RSOV is not in Spain (which the 20 do not include). The library guide's page metadata reads datePublished 2025-11-11, dateModified 2026-10-09, the same modification date as the what's-new page, so neither is the later word. The what's-new launch page *Drive top-of-search visibility and engagement by reserving branded share of voice through Sponsored Brands* (advertising.amazon.com/resources/whats-new/sponsored-brands-reserve-share-of-voice, published 2025-10-22, dateModified 2026-10-09, read 2026-11-09) also lists 20, but with Egypt in place of Ireland: US, CA, MX, BR, DE, FR, IT, UK, BE, PL, NL, SE, TR, EG, SA, AE, AU, JP, IN, SG; it names sellers, vendors, partners and agencies as users, through the console and the API. [official] See Open questions 1–3.
+
+6. **GUIDANCE — what the reservation delivers.** Goals page: "We strive to deliver 95% SOV at the campaign level…"; top-of-search impression share for a reserved keyword is "the amount of impressions the advertiser received for the reserved keyword divided by the total amount of impressions across all ads for the reserved keyword", with the worked example Kitchen Smart, 600 of 1,000, = 60%. Library guide FAQ: it "guarantees coverage… at a fixed cost per mille (CPM)"; formats are "Sponsored Brands video, store spotlight, and collections" (the help page's builder steps name only Product collection and Video). A further RSOV-only format, Brand Gallery — "a media block, brand messaging, a call-to-action linking to the Brand Store, and a carousel of three to five Brand Store categories or collections" — is announced on *Drive branded search engagement with Sponsored Brands Brand Gallery* (advertising.amazon.com/resources/whats-new/sponsored-brands-brand-gallery, published 2026-05-28, read 2026-11-09; beta "with 4 advertisers, March – May, 2026"). The 2025-10-15 release note and the launch page say the ad will appear "most of the time". [official] "Strive" and "most of the time" are not a contractual floor on the help pages read; the help pages point to separate "additional terms that apply to Reserve Share of Voice campaigns", which this seat did not read.
+
+7. **GUIDANCE — Amazon's beta results (human readers only; Amazon internal, not independent).** Library guide: beta with 15+ advertisers, May–August 2025: Sponsored Brands top-of-search impression share on reserved keywords rose from 62.7% to 99.3%, CTR from 3.5% to 4.1%, click-attributed sales +143%, lost top-of-search sales from 10.3% to 0.3%. Samsung case study (advertising.amazon.com/library/case-studies/sponsored-brands-reserve-sov-samsung, read 2026-11-09): SB top-of-search impression share 44.9% → 99.8%, click-attributed sales +187%, conversion rate 13.2% → 14.3%, lost top-of-search sales 7.4% → 0.04%, footnoted "Amazon Marketing Cloud, US, 2025… a single advertising campaign for an early-stage or newly launched product… Results may not be typical". Bayer case study (…/sponsored-brands-reserve-sov-bayer, read 2026-11-09): eight OTC brands in Germany, SB top-of-search impression share 77% → 97%, 7.8× impressions, 5.6× click-attributed top-of-search sales; the Bayer quote reads "Sponsored Brands reserve share of voice broke the CPC auction spiral exactly where we can't afford to lose visibility". [official] These are before/after comparisons of attributed sales, not holdout tests of incrementality.
+
+8. **RULE — the API lifecycle of a reservation.** *Reserve share of voice* API guide (guides/sponsored-brands/reserve-share-of-voice.md, read 2026-11-09) and the merged v1 model (read 2026-11-09): keyword recommendations `POST /adsApi/v1/create/recommendations/sb`; reservability check `POST /adsApi/v1/create/keywordReservationValidations/sb` (up to 1,000 keywords per call, `SBCreateKeywordReservationValidationRequest` maxItems 1000; each answer carries `isReservable`, a required `keywordReservationValidationId` and, when false, `reservationRejectedReason`, e.g. "No valid brand or trademark was found in the keyword."); pricing `POST /adsApi/v1/create/brandedKeywordsPricings/sb` (returns `keywordsPricing` and `rejectedKeywords`; `advertisingDealId` is omitted when pricing a new deal); the deal `POST /adsApi/v1/create/advertisingDeals/sb`; deal targets `POST /adsApi/v1/create/advertisingDealTargets/sb` (at least five), `…/query/…` and `…/delete/…`; `POST /adsApi/v1/update/advertisingDeals/sb` moving the state from `DRAFT` to `PROPOSED` (the only two states in `SBAdvertisingDealState`), which triggers pricing; `POST /adsApi/v1/query/advertisingDeals/sb`; the campaign itself on `POST /sb/v4/campaigns` with `targetPGDealId`; and `POST /adsApi/v1/delete/advertisingDeals/sb`. "A deal in PROPOSED state cannot be modified"; a deal can be deleted only before its start time minus 12 hours, and deletion archives its campaigns; only the day part of start and end is used, in the marketplace's time zone. [official] Detail in `method/ads-api-surface.md`.
+
+### C. Grow brand impression share: buying top-of-search share in the auction
+
+1. **RULE — the goal's shape.** "Grow brand impression share" is the vCPM Sponsored Brands goal that serves top-of-search (goals page GXXPFXLSVX9VD3U9; SB v4 `BRAND_IMPRESSION_SHARE`). Formats on top-of-search require a brand Store (goals page table). The goal predates the 2024-06-13 release note that lists it among SB v4 features: the 2023-11-06 note already calls "grow brand impression share" campaigns, "transacted on `vCPM` basis", "recently launched", and the 2023-12-26 note adds video-only creatives under it (release-notes/index.md, read 2026-11-09). [official]
+
+2. **GUIDANCE — bid recommendations come from history.** Goals page: bid recommendations are "built from historical winning bids…". [official]
+
+3. **RULE — the v1 KPI enum carries top-of-search share as a target.** The merged v1 model's `KPI` enum includes `TOP_OF_SEARCH_IMPRESSION_SHARE` beside `NEW_TO_BRAND_PURCHASES`, and the `BidStrategy` enum includes `NEW_TO_BRAND` ("Optimizes bidding to maximize new-to-brand customer acquisitions") and `PRIORITIZE_KPI_TARGET` (read 2026-11-09). [official] Which ad products accept which KPI is in `beyond/goals-and-economics.md` and `beyond/automation-and-amazon-tools.md`.
+
+4. **RULE — budget rules can key on impression share.** The 2023-04-06 release note: Sponsored Brands performance-based budget rules can use "top-of-search impression share (`metricName` is `IS`)" and "percent of sales new-to-brand (`metricName` is `NTB`)" (release-notes/index.md, read 2026-11-09). [official] Detail in `areas/budgets-and-pacing.md`.
+
+### D. Share metrics: what Amazon publishes about a seller's position against others
+
+Amazon publishes five share measures. None names a competitor.
+
+1. **RULE — search term impression share and rank are account-wide, per search term.** *Downloadable reports in sponsored ads* (help GBYSPTSLR337JMLH, lastModified 2026-08-19, read 2026-11-09): the Search Term Impression Share report gives "for each search term, the percentage share and numeric rank of your impression share of the total impressions relative to other advertisers", for Sponsored Products and Sponsored Brands. The v1 metric pages (guides/reporting/ads-v1/metrics/impression-share/search.md, read 2026-11-09): `metric.impressionShare` and `metric.impressionShareRank` ("numeric rank of your account-wide impression share compared to all other advertisers") require `adProduct.value`, `advertiserAccount.id` and `searchTerm.value`, and are **incompatible** with campaign, ad group, ad, advertised product, target and placement classification; v3 names "Search Term Impression Share" and "Search Term Impression Rank". The what's-new page *Search term impression share report for Sponsored Products* (March 22, 2021; …/whats-new/search-term-impression-report-sponsored-products, read 2026-11-09) gives the reading within one ad product: a 20% share "means that you won 20% of all Sponsored Products ad impressions for that search term on that date"; rank 3 means "you received the third most Sponsored Products ad impressions". [official] So the denominator is the ad impressions of one ad product on that term, not every ad on the page. Perpetua's explainer (perpetua.io/blog-amazon-search-term-impression-share-rank-report/, read 2026-11-09) adds a working note: "Impression share is account-wide; however, search terms and impressions appear separately for each ad group/campaign in the report", so impressions for a term have to be summed across rows. [opinion, Perpetua]
+
+2. **RULE — top-of-search impression share has the seller's own eligibility as its denominator.** v1 `metric.topOfSearchImpressionShare`: "percentage of impressions that appeared in top search results compared to the number of times they were eligible"; compatible with campaign; incompatible with ad group, target, search term and placement classification; v3 "Top-of-search Impression Share" (same metric page, read 2026-11-09). The 2021-09-22 what's-new page *Top-of-search impression share (top-of-search IS) to measure brand awareness* (…/whats-new/top-of-search-impression-share-metric, read 2026-11-09) frames it as share of the top-of-search impressions the campaign was "eligible to serve on", with a 90-day lookback, and suggests one can "adjust bids by placement". [official] In v3, the columns page's entry reads "**Report types**: spCampaigns, sbCampaigns, sbTargeting"; the *Targeting reports* page lists it in the base metrics of both the Sponsored Products and the Sponsored Brands targeting reports; and the *Campaign reports* page lists it under Sponsored Products "Group by `campaign`" and "Group by `campaignPlacement`" and under Sponsored Brands "Group by `campaign`" (guides/reporting/v3/columns.md, report-types/targeting.md and report-types/campaign.md, all re-read 2026-11-09), while v1 holds it at campaign level only (`areas/bids-and-placements.md` §J). So the column is documented for SB as well as SP in v3; a list naming SP alone is incomplete against these three pages. The RSOV definition (§B.6) is different again: there the denominator is "all ads for the reserved keyword". [seat] Three Amazon metrics called "impression share" have three different denominators: one ad product's impressions on a search term across all advertisers (STIS), the campaign's own eligible top-of-search impressions (TOS IS), and all ads in a reserved keyword's Sponsored Brands top slot (RSOV).
+
+3. **GUIDANCE — the Brand Impression Share dashboard is a console view, brand-level, relative to peers.** *Brand impression share* (help GNU4AUAUHS76DNDR, lastModified 2026-02-25, read 2026-11-09): in Insights and Planning → Brand impression share, it shows "top-of-search impression share on their brand terms across their Sponsored Products and Sponsored Brands campaigns over time"; it identifies brand terms automatically, rolls up across all advertising accounts of the brand, and shows the share "relative to peers"; a second section lists campaigns and keywords with filters such as "raising bids and budgets for terms with low impression share but high impression volume". Note on the page: "It represents the advertising share of voice that you have on brand terms." [official] The RSOV library guide places the "Lost Top-of-Search Sales metric… in Amazon Ads Agent under the Insights and Planning tab within the Brand Impression Share report". [official] The seat found no API operation that returns this dashboard (Open question 4).
+
+4. **RULE — keyword recommendations carry share and rank.** Sponsored Products keyword recommendations version 4 add "search term impression share" ("percentage share of all ad-attributed impressions you received on that keyword in the last 30 days") and rank, as `searchTermImpressionShare` and `searchTermImpressionRank`, on `POST /sp/targets/keywords/recommendations` (guides/sponsored-products/keywords/suggestions.md, read 2026-11-09); the same guide says version 5 "includes all the features of version 4" and asks integrators on version 3 or 4 to move to version 5 "as soon as possible". Sponsored Brands keyword recommendations gained the same pair on 2023-05-25, and the SB campaign insights API on 2023-03-07 (release-notes/index.md). [official]
+
+5. **RULE — the partner share-of-voice insight is gated and brand-level.** *Share of voice insights* (guides/recommendations/partner-opportunities/catalog/share-of-voice-insights.md, read 2026-11-09), for Amazon Ads Partner Network members only: share of voice "is the percentage of impressions achieved for a brand and browse node as a result of any keyword compared to total Sponsored Products impressions for the same brand or similar brands"; four opportunity types: brands losing or gaining share of voice, in Sponsored Products and in Sponsored Brands. The Partner Opportunities overview (guides/recommendations/partner-opportunities/overview.md, read 2026-11-09) describes the API as giving "insights to Partner Network members for all of the partner's linked clients" and requires being "a registered member of the Amazon Ads Partner Network". [official] A seller's own connection is not a partner (Open question 5).
+
+6. **RULE — Search Query Performance gives share across the funnel.** See §G.3: SQP's `asinImpressionShare`, `asinClickShare`, `asinCartAddShare` and `asinPurchaseShare` fields are the requested ASINs' share of a query's totals from the search results page; that these combine organic and paid is Trellis's reading [opinion], not a sentence on the Amazon pages read.
+
+7. **GUIDANCE — category benchmarks show peers' quartiles, not peers' names.** The Category Benchmark report gives the "median, the bottom 25%, and the top 25% values achieved by your peers (by retail performance)"; the help page's table marks it for Sponsored Brands only (help GBYSPTSLR337JMLH). The v3 `crossProgramBenchmarks` report (guides/reporting/v3/report-types/benchmarks.md, read 2026-11-09) needs the caller to be "a brand owner or representative on Amazon Stores"; groups by `brandCategoryBenchmarks`; carries brand, browseCategory, adProduct, adFormat, campaignGoal and `peerSetSize`; and gives P25/P50/P75 for percentOfPurchasesNewToBrand, newToBrandPurchaseRate, costPerNewToBrandPurchase, ctr, cpc, cpm and video rates; data from January 1, 2025, refreshed every 48 hours, ranges 456 days (monthly or weekly) or 90 days (daily). v1 dimensions are `benchmarkBrand.id`/`.name`, `benchmarkBrowseCategory`, `benchmarkBrowseCategoryMarketplace` and `benchmarkCampaignGoal` (…/ads-v1/report-types/cross-program-benchmarks.md); the benchmark-brand dimension page defines the brand as "associated with a campaign used for benchmark comparisons derived from the products promoted in the campaign" (…/ads-v1/dimensions/benchmarks/benchmark-brand.md, read 2026-11-09). The 2026-09-25 release note adds P25/P50/P75 benchmark metrics to unified reporting (e.g. `metric.addToCartRateP25`); 2026-05-18 made benchmarks generally available in 18 marketplaces. [official] Figures human readers only.
+
+### E. Defending one's own detail pages
+
+1. **RULE — a seller can target its own ASINs.** *Set up product targeting* (help GB2JECV9CJK6R6AL, lastModified 2026-02-20, read 2026-11-09) lets an advertiser "search for specific products" and add them, with no restriction on whose; the negative-targeting help page (§E.6) names "your own or a competitor" detail pages explicitly, and Amazon's best-practices guide (§E.2) recommends targeting one's own. [official]
+
+2. **GUIDANCE — Amazon recommends it.** *Best practices for your Sponsored Products ads* (advertising.amazon.com/library/guides/sponsored-products-best-practices, read 2026-11-09): "Target your own product pages to showcase your product range : Targeting your own products and bidding higher on them can help you increase visibility for your range on your product pages." [official]
+
+3. **RULE — expanded product targeting is on by default and widens a target beyond the named ASIN.** Same help GB2JECV9CJK6R6AL: expanded product targeting "is the default setting", with the "Expanded box… automatically selected"; it chooses "closely related products based on past interactions and views", updates daily, and shows expanded matches with a "Product-expanded" prefix in reports; negatives can exclude products or brands from the expanded list. Category targets can be refined by brand, price and rating, by Prime for SP and SD, by age range in Toys and by genre in Books. [official] [seat] An own-ASIN defence target left on its default therefore also serves on pages Amazon judges close to the seller's ASIN, which may be competitors' pages; the report prefix is how a reader separates the two (Open question 10).
+
+4. **RULE — Sponsored Brands product targets: same-as or expanded.** Sponsored Brands product targeting uses `ASIN_SAME_AS` or `ASIN_EXPANDED_FROM`, one ASIN per clause, and `ASIN_CATEGORY_SAME_AS` for categories (guides/sponsored-brands/targeting/product-targeting.md, read 2026-11-09). The same guide notes: "In Sponsored Brands and Sponsored Display campaigns… we may increase your bid up to 300%… a bid of $1 up to $4", and adds: "For Sponsored Brands, these bid adjustments are available only on campaigns where you select 'Drive page visits'" as the goal. [official] Figures human readers only.
+
+5. **RULE — Sponsored Display reaches detail pages and audiences of a brand.** *Understand targeting in Display campaigns* (help GPKPWTLW6B8T683N, lastModified 2026-03-19, read 2026-11-09) includes "prospecting for new customers who viewed the detail pages of or purchased products similar to yours". The SD audience guide (guides/sponsored-display/audience-targeting.md, read 2026-11-09) lists `views(exactProduct …)`, `views(similarProduct …)` (lookbacks 7–90 days), `purchases(exactProduct …)` and `purchases(relatedProduct …)` (7–365 days), and the refinement `asinBrandSameAs=<brand id>`: "Target an audience that has viewed or purchased products in the given brand in the past". [official] [seat] With a competitor's brand id, that refinement is an audience-conquest tool; with one's own, a remarketing one.
+
+6. **RULE — negative product targeting keeps a seller's ads off named pages, own or competitor.** *Add negative keywords or negative products targeting* (help G8W49VU65XQ4T2NS, lastModified 2026-10-20, read 2026-11-09): "To keep your Sponsored Products ads off specific product detail pages, whether your own or a competitor, use negative product targeting." Ad-group negatives affect the product and auto targets of that ad group; campaign negatives affect every ad group. The console fields are "Exclude brands" and "Exclude products". Changes take effect after a buffer: 72 hours for negative keywords, 96 hours for negative products. Negative phrase up to 4 words, negative exact up to 10, 80 characters. [official] Figures human readers only.
+
+7. **RULE, conflicting — who may add campaign-level negative products to manual campaigns.** The API guide (guides/sponsored-products/negative-targeting/product-brand.md, read 2026-11-09): "Both auto and manual Sponsored Products campaigns can use negative product targeting… exclude certain products and brands", created on `POST /adsApi/v1/create/targets` (answers 207), with the warning "Campaign-level negative product targeting for manual campaigns is only available for vendors. Sellers can add campaign-level negative targeting to auto campaigns, but not manual." `areas/search-terms-and-negatives.md` records a 2025-05-22 release note saying every campaign may carry them (US), and an SP 3.0 model allowing AUTO only. [official] See Open question 6.
+
+8. **RULE — Sponsored Brands negative product targets are ad-group level.** `POST /sb/negativeTargets` with `adGroupId`, `campaignId` and `expressions.type` `asinBrandSameAs` or `asinSameAs`; brand ids come from `POST /sb/recommendations/targets/brand` (guides/sponsored-brands/targeting/negative-product-targeting.md, read 2026-11-09). [official] Which model carries which SB negative path (re-parsed 2026-11-09): the YAML model titled "Amazon Ads API - Sponsored Brands", version "3.0" (`openapi/en-us/sponsored-brands/3-0/openapi.yaml`), carries `/sb/negativeKeywords`, `/sb/negativeKeywords/{keywordId}`, `/sb/negativeTargets`, `/sb/negativeTargets/list`, `/sb/negativeTargets/{negativeTargetId}` and `/sb/recommendations/targets/brand`, and no `/sb/negativeTargets/brands/recommendations`; the JSON file titled "Sponsored Brands 3.0" (`d1y2lf8k3vrkfu.cloudfront.net/openapi/en-us/dest/SponsoredBrands_prod_3p.json`, 50 paths) and the SB 4.0 model ("Sponsored Brands campaign management 4.0", 48 paths) both carry `GET /sb/negativeTargets/brands/recommendations` and neither carries `/sb/negativeKeywords` or `/sb/negativeTargets`. The *Sponsored Brands feature availability by marketplace* guide links `/sb/negativeKeywords` and `/sb/negativeTargets` to `sponsored-brands/3-0/openapi`. [official] So two different files are both called "Sponsored Brands 3.0"; `method/ads-api-surface.md` owns the surface and `areas/search-terms-and-negatives.md` carries the three-model table.
+
+8a. **GUIDANCE — Amazon's own API description on targeting one's own brand in Sponsored Brands.** The SB v4 OpenAPI operation `GET /sb/negativeTargets/brands/recommendations` ("Get brand recommendations for negative targeting"; `openapi/en-us/sponsored-brands/4-0/openapi.json`, read 2026-11-09; the same operation, `operationId` `SBTargetingGetNegativeBrands`, is in the JSON "Sponsored Brands 3.0" file and not in the SB 3.0 YAML) reads: "Returns brands recommended for negative targeting. Only available for Sellers and Vendors. These recommendations include your own brands because targeting your own brands usually results in lower performance than targeting competitors' brands." Marketplaces: US, CA, MX, UK, DE, FR, ES, IT, IN, JP. [official] The sentence gives no measure of "performance" and no data. [seat] It sits beside Amazon's SP guidance to target one's own product pages (§E.2) and its "Protect" strategy (§A.1); the two are not contradictory on their face (one is about SB product targets on one's own brand, the other about SP visibility on one's own pages), but a reader weighing an own-brand SB product target meets Amazon on both sides.
+
+9. **GUIDANCE, conflicting — negative match types.** *What is keyword targeting?* mentions negatives as "exact, phrase, or broad"; the help page G8W49VU65XQ4T2NS lists negative phrase and negative exact only. [official] Open question 7.
+
+10. **RULE — which report shows the pages a product target served on, by ad product.** Help GBYSPTSLR337JMLH: the matched target report lists "ASINs, and their related product detail pages, your ads appeared against and resulted in at least 1 click" — and the page's report table marks it for **Display only**, not Sponsored Products or Sponsored Brands; the v1 dimension `matchedTarget.value` ("The product page where the ad appeared", v3 name `matchedTargetAsin`) likewise lists Sponsored Display as its only ad product (guides/reporting/ads-v1/dimensions/targeting/matched-target.md, read 2026-11-09). The purchased product report "details on the non-advertised SKUs (ASINs for SD) a customer purchased after clicking your ad" (Sponsored Products and Display). For Sponsored Products and Sponsored Brands, the v3 search term report "only include[s] impressions that resulted in at least one ad click", and "If a placement does not have a search keyword associated with it on a product detail page, the search term in the report will be an asterisk `*`" (guides/reporting/v3/report-types/search-term.md, read 2026-11-09). [official] [seat, not on an Amazon page read] Practitioners read ASIN-valued rows of the SP search term report as the detail pages a product target matched; no Amazon page read states this. Either way only clicked pages appear, so impressions on pages that drew no click stay unseen.
+
+### F. Conquesting and the policy line between a keyword and creative
+
+1. **RULE — no Amazon page read forbids targeting another brand's name as a keyword or another brand's ASIN as a target.** The *Targeting* policy page (help GD6SPCK5RJZL73PU, lastModified 2026-08-03, read 2026-11-09) has no competitor or trademark rule; the ad-policy change log (help G8X8M375UPCN4UUC, lastModified 2026-04-27, read 2026-11-09) has entries from October 2025 through April 2026 and none on competitors or trademarks. Amazon's targeting guide names "Conquest : target competitors' products" as a strategy (§A.1), and *What is keyword targeting?* suggests targeting "products from other brands where your offer adds strong value, such as when other brands' products have lower star ratings, fewer reviews, and higher price points." [official] The one place Amazon refuses another company's trademark as a keyword is reserve share of voice (§B.2). This matches `areas/keyword-and-product-targeting.md` §P, re-read for this file.
+
+2. **RULE — creative may not name or disparage a competitor.** *Prohibited content and products for Amazon O&O supply* (help GTEEH3E3J2BDELBA, lastModified 2026-04-10, read 2026-11-09) prohibits "Disparaging or criticizing a named competitor". The Sponsored Brands and display moderation guide (advertising.amazon.com/library/guides/sponsored-brands-display-ads-moderation, re-read 2026-11-09): "Don't use competitive language"; "Ads on Amazon cannot explicitly call out your competitors by name. You may include references to a 'leading brand' or other generic phrasing."; "Don't use superiority claims (the best, the biggest, better than ) or mention competitors by name." Superiority claims "will be rejected unless they can be substantiated". [official]
+
+3. **RULE — comparative claims need substantiation; for health products, in the ad itself.** *Claims and substantiation* (help G6UU66YLSCUYDULQ, lastModified 2026-01-28, read 2026-11-09), for healthcare products "(excluding personal care products)": "Comparative claims: Claims that compare a product or service to a competitor's product (for example, 'Provides 50% more pain relief than the leading brand'). If the claim is on the product packaging, no further substantiation is needed. Otherwise, substantiation, including source and date (within 24 months), is required in the ad creative, on product packaging, or landing page." [official]
+
+4. **RULE — no competitive pricing claims.** *Pricing and Promotions* (help GKP5TCJWVN5W7AFU, lastModified 2026-03-26, read 2026-11-09): ads "must not contain superlatives relating to pricing… or contain pricing claims that are competitive, misleading, or can't be proven", with examples "Competitive price" and "Unbeatable price". [official]
+
+5. **RULE — rights to every element of the creative.** *General requirements* (help G4U9JBZH3664JWX2, lastModified 2026-07-22, read 2026-11-09): advertisers must "Obtain and maintain all necessary rights and permissions for all content in ads, such as logos…", follow the Brand Usage Guidelines for Amazon marks, and "must not use unauthorized content that uses 'Amazon's Choice,' 'Best Seller'". The O&O page also says "Brand Stores can't be titled 'Sale,' 'Discount,' 'Deals,' or other terms that imply a promotion." [official]
+
+6. **Mechanism — why a keyword is not creative.** [seat] In Sponsored Products the shopper sees the advertised product's own title, image, price and rating; the keyword that matched is never shown. In Sponsored Brands and Display the advertiser writes a headline and supplies a logo, which moderation reads. The policy line Amazon draws follows that split: targeting by another brand's name is unrestricted outside RSOV; naming that brand in what the shopper reads is forbidden.
+
+7. **[legal] — the US case law around Amazon's own use of a competitor's mark in search.** *Multi Time Machine v. Amazon.com* (9th Cir., No. 13-55575, decided October 21, 2015; read through the FindLaw case summary, 2026-11-09, which affirms summary judgment for Amazon): the search "mtm special ops" on Amazon returned "a list of other watches not sold by Multi Time Machine"; the court held that "no reasonably prudent customer accustomed to shopping online would likely confuse the source of the products", because Amazon labels each result with its name, manufacturer and photo. *1-800 Contacts v. FTC* (2d Cir., June 11, 2021; read through the AIPLA summary): the FTC had found anticompetitive 1-800 Contacts' settlements requiring rivals to "refrain from engaging in search advertising using specific keywords" and to "employ 1-800 Contacts' trademarks as negative keywords"; the Second Circuit vacated the order under the rule of reason, saying "we owe significant deference to arm's length use agreements" (FTC matter page, ftc.gov/legal-library/browse/cases-proceedings/141-0200-1-800-contacts-inc-matter: agreements with "at least 14 competing online contact lens retailers", ALJ initial decision October 20, 2017, Commission opinion and final order 2018-11-14, FTC docket 9372; 18-3848 is the Second Circuit appeal number; the page's last entry is the FTC's petition for rehearing of August 6, 2021). In the EU, *Google France* (CJEU C-236/08, 2010) is the leading keyword case; EUR-Lex did not answer and it is cited here only through Bechtold & Tucker (below). [legal] These are US and EU reference points, not advice; trademark law is by jurisdiction (Open question 13).
+
+### G. Brand Analytics as a competitive signal
+
+1. **RULE — who can see it.** sell.amazon.com *Amazon Brand Analytics* (sell.amazon.com/tools/amazon-brand-analytics, read 2026-11-09): a Professional selling account and the Brand Representative role in Brand Registry. Top Search Terms "lists… the top three products, categories, and brands associated with each keyword"; Search Query Performance "lists keywords customers use to find your brand"; the page also lists Market Basket, Repeat Purchase, Demographics and Customer Loyalty Analytics. SP-API's analytics report-type page (developer-docs.amazon/sp-api/docs/report-type-values-analytics, read 2026-11-09): `GET_BRAND_ANALYTICS_SEARCH_TERMS_REPORT` needs the Brand Analytics role, Brand Registry and brand representative; `GET_BRAND_ANALYTICS_SEARCH_QUERY_PERFORMANCE_REPORT` and `GET_BRAND_ANALYTICS_SEARCH_CATALOG_PERFORMANCE_REPORT` are for "Sellers who have the Brand Analytics Selling Partner API role and are registered in Amazon Brand Registry", cover 18 marketplaces, are request-only and return JSON; the Search Terms report is "limited to approximately 12 million items". [official]
+
+2. **RULE — Top Search Terms is the one Amazon source that names competitors' products on a query.** `sellingPartnerSearchTermsReport.json` (amzn/selling-partner-api-models, schemas/reports, read 2026-11-09): "top-3 clicked ASINs by search keyword and department"; periods DAY, WEEK, MONTH, QUARTER; fields `departmentName`, `searchTerm`, `searchFrequencyRank`, `clickedAsin`, `clickShareRank`, `clickShare` ("fraction of clicks") and `conversionShare`. [official] [seat] For a seller's own brand query, the three clicked ASINs and their click shares show whether a rival's product is among the top three clicked on that query, which is the nearest Amazon-sourced read of "who is taking my brand traffic" (SB brand recommendations by keyword, §E.8, name brands but carry no share). It reports clicks regardless of paid or organic, and only the top three.
+
+3. **RULE — Search Query Performance gives the requested ASINs' share at four stages.** `sellingPartnerSearchQueryPerformanceReport.json` (read 2026-11-09) and the SP-API report-type page: periods WEEK, MONTH or QUARTER, Sunday–Saturday weeks, one marketplace, and a **required** `asin` report option — "A space-separated list of ASINs for which you can request the report. There is a 200 character limit." — so the API report is ASIN-scoped, not a brand view; fields `searchQuery`, `searchQueryScore`, `searchQueryVolume`; impressions `totalQueryImpressionCount`, `asinImpressionCount`, `asinImpressionShare`; clicks `totalClickCount`, `asinClickShare`, with shipping-speed counts; cart adds with `asinCartAddShare`; purchases with `asinPurchaseShare`. `totalQueryImpressionCount` is "The total number of impressions that originate from the search results page for all ASINs with the current query." [official] The Seller Central help page for SQP (G8J4CB5ZBF3NX7TP) serves only a JavaScript shell to a plain GET and was not read (Unreachable). Trellis describes a brand view and an ASIN view in the console, and that "counts combine organic and paid" and cover only queries on which the brand already appeared [opinion, Trellis]. Detail of SQP as a listing tool is in the listing-playbook research; this file uses only its share columns.
+
+### H. Stores, Store Spotlight and Brand Store insights
+
+1. **RULE — Store Spotlight needs a Store with at least three subpages.** guides/sponsored-brands/ads/store-spotlight.md (read 2026-11-09): created on `/sb/v4/ads/storeSpotlight`; "your Store on Amazon with at least three subpages", each given with a page title, an ASIN present on the subpage and a URL. The help page *Understand Sponsored Brands* (GGWFYHL27MFXLHS6, lastModified 2026-10-20, read 2026-11-09) words it as a Store that "must have at least 4 pages, each with 1 or more unique products", and *Your complete guide to Sponsored Brands* as "at least three sub-pages to be eligible, with a unique product on each sub-page". [official] [seat] A home page plus three subpages is four pages, so the three wordings can be read as one rule. §L carries Store Spotlight as a defence and conquest format.
+
+2. **RULE — Store insights by API.** guides/brand-store/overview.md and insight-metrics.md (read 2026-11-09): the overview guide still documents `GET /v2/stores` as the call that returns `brandEntityId` ("The `brandEntityId` is required in order to use the Brand Stores metrics APIs"), but the *Deprecations* page (release-notes/deprecations.md, re-read 2026-11-09) lists "V2 Stores | `/v2/stores` | /brand/stores/v1/storePages/list | January 2026 | June 1, 2026", so on 2026-11-09 `/v2/stores` is past its shutoff. After it, `brandEntityId` is in `POST /brand/stores/v1/stores/list` ("Lists all Stores for Advertiser"; `StoreInfo` carries `brandEntityId`, `storeId`, `storeName`, `storeStatus`; `maxResults` 1–30; *Brand Home* model, `BrandHome_prod_3p.json`, read 2026-11-09) and in SB `GET /brands` (`brandEntityId`: "The Brand entity identifier available exclusively for seller advertisers"; SB 3.0 YAML). `POST /brand/stores/v1/storePages/list`, the named replacement, lists a Store's pages (`storePageId`, `storePageName`, `storePageStatus`, `storePageUrl`) for an identifier that is "store's brand-/sub-entityId and storeId". [official] `method/ads-api-surface.md` owns these paths. Insights: `POST /stores/{brandEntityId}/insights` over at most 100 days; metrics VIEWS, VISITS, VISITORS, ORDERS/UNITS/SALES ("within 14 days" of a visit), DWELL, `PEER_DWELL`, DWELL_TIME, BOUNCE_RATE, NEW_TO_STORE, SCORE_LEVEL and RECOMMENDATIONS, by the dimension SOURCE. The merged v1 model adds `/adsApi/v1/query/brandStores`, `/adsApi/v1/query/brandStorePages` and Store edition publishing paths. [official] `PEER_DWELL` is the only peer comparison on a Store.
+
+3. **RULE — Brand Metrics compares a brand to category peers.** guides/reporting/brand-metrics/overview.md and how-to.md (read 2026-11-09): open beta for Brand Registry professional sellers and vendors in US, CA, MX, UK, DE, FR, IN, IT, ES, JP, AU; Awareness and Consideration indices "compare your performance to peers"; `POST /insights/brandMetrics/report`, then `GET /insights/brandMetrics/report/{reportId}` until `SUCCESSFUL`, then `location`; columns include `customerConversionRate` with category median and top-performer values. [official]
+
+4. **RULE — Brand halo in reports.** v3 columns (guides/reporting/v3/columns.md, read 2026-11-09): `attributionType` "Describes whether a purchase is attributed to a promoted product or brand-halo effect" (`sbPurchasedProduct`); `asinBrandHalo` "Represents a product that was purchased after an ad click but has a different `SKU`/`ASIN` than what was advertised" (`sdPurchasedProduct`); and halo variants of add-to-cart and add-to-list (DSP report types, e.g. `addToCartBrandHalo` in `dspProduct`). [official] Detail in `beyond/metrics-and-attribution.md`.
+
+### I. The auction, reserve prices and the 2026 dispute over them
+
+The auction rules decide what a bidding war costs, so the current state of the dispute belongs here. Mechanics of bidding are in `areas/bids-and-placements.md`.
+
+1. **GUIDANCE — Amazon's description of its auction.** *Understand Amazon Ads auctions* (help GEXGZNBBWPZZK7P8, lastModified 2026-04-27, read 2026-11-09): relevance factors decide winners as well as bids, and "ads with lower bid amounts, but higher relevance… can and do win", over "billions of auctions each day". On reserves: "Some reserves help allocate ad space by setting a bid threshold… Other reserves influence your final CPC or vCPM based on factors including the likelihood that the ad will lead to a sale, predicted return on your ad spend, the value of any competing bids, placement, context… As a result, the price charged to you may exceed the runner-up bid but will never exceed the maximum bid you authorize." And: "Rather than trying to optimize for specific auctions, focus on setting bids that align with your overall goals." [official]
+
+2. **[official, regulator — allegations, untested] — the FTC complaint.** Press release *FTC, States Sue Amazon over Secret Ad Surcharge Scheme*, 2026-08-31 (ftc.gov/news-events/news/press-releases/2026/08/ftc-states-sue-amazon-over-secret-ad-surcharge-scheme, read 2026-11-09): the FTC (vote 2-0) and 22 states sued in the U.S. District Court for the Western District of Washington. Allegations: Amazon told advertisers winners pay "one cent more than the next highest bidder"; since 2019 a "soft reserve price", described internally as a "proxy 2nd price that we calculate", acted as an "invented auction participant"; Sponsored Products advertisers paid "their own winning bid close to 80% of the time"; Amazon "effectively converted its nominally 'second price' auction into a first price auction". Scope: Sponsored Products, Sponsored Brands and "Display Ads". Update, 2026-11-05 (…/2026/11/new-details-emerge-amazon-ads-auction-rigging-case, read 2026-11-09): alleged surcharges up to 50% on ordinary days and up to 250% on Prime Day and Black Friday; 2021 holiday CPC +20% with some advertisers over 200%; a 2022 holiday figure of $87M; a 2023 Prime Day food advertiser over 100%; Amazon's internal assessment of that 2023 case quoted as "destroying campaign performance"; an account manager quoted: "we were informed there is not a bug, or a change in auction style, so no worries there" and "This year the marketplace is even more crowded than any year previous". Figures human readers only; these are allegations in a pending case.
+
+3. **[official, Amazon's response] — Amazon's answer.** *Amazon's response to the FTC's lawsuit regarding Sponsored Ads* (aboutamazon.com/company-news/amazon-ftc-sponsored-ads-lawsuit-response, published 2026-08-31, read 2026-11-09): "The case centers on generalized second price auction dynamics"; "In no scenario does an advertiser pay more than their bid"; a hard reserve is "the minimum a bid must surpass to enter an auction"; a soft reserve is "a real-time minimum value that seeks to better reflect what each placement is actually worth". The mechanism in Amazon's own words: "When the winning advertiser's bid exceeds both the hard and soft reserve, they pay the soft reserve, which is less than they were willing to pay. When the winning advertiser's bid exceeds the hard reserve but doesn't meet the soft reserve, we still grant the placement to that advertiser and they pay their bid." Figures: "Average winning bids fell 50% from 2019 to 2025 on Sponsored Products search ads"; "in 2024, approximately 92% of selected Sponsored Products ads were not the highest bid"; "The mean winning advertiser's bid is typically about the 128th bid by amount"; average SP search CPC "remained flat adjusted for inflation" 2019–2024 while "conversion rates grew 24% from 2021 to 2025"; advertisers "saved over $8 billion from 2021 to 2025"; the suit is "misguided". Figures human readers only.
+
+4. **[seat] — what the dispute changes for a bidding war.** Both parties agree that a soft reserve exists, that a winner can pay more than the runner-up, and (in Amazon's own description) that a winner whose bid falls between the hard and the soft reserve pays its own bid; they disagree on how often that happens and on whether it was disclosed. Under either account, raising a max bid in a contested brand auction can raise the price paid even when the rival does not move, because the price is not set by the rival's bid alone. That makes "bid one cent above the competitor" an unreliable model of cost on Amazon.
+
+### J. New surfaces where brand competition now happens
+
+1. **RULE — prompts and AI surfaces.** The v3 prompt-ad-extension reports (`spPromptAdExtension`, `sbPromptAdExtension`; guides/reporting/v3/report-types/prompt-ad-extension.md, read 2026-11-09): "Prompts are a new ad format that integrates into your existing Sponsored Products and Sponsored Brands campaigns with zero additional setup"; 90-day ranges, 95-day retention, `marketplaceId` US only; "Prompts with clicks will show in your existing Sponsored Products or Sponsored Brands reporting, and you can pause individual prompts through the Amazon Ads console." The launch page *Sponsored Products prompts and Sponsored Brands prompts* (advertising.amazon.com/resources/whats-new/unboxed-2025-sponsored-products-and-sponsored-brands-prompts, published 2026-03-10, read 2026-11-09): prompts moved "from open beta to general availability in the U.S. on March 25, 2026", campaigns are "automatically enrolled", "Prompts can appear in shopping results and product detail pages on Amazon", and "we will begin to charge for these ads as part of your CPC bidding and billing parameters". The 2026-08-24 release note adds portfolio, currency and conversion-rate columns to `spPromptAdExtension` and `spVideoAdExtension` (Sponsored Products). The 2024-09-05 note on Rufus placements: "Your campaign reports will not include Rufus metrics". The 2025-07-15 note, titled "Alexa+ tests product discovery through conversational AI", says Sponsored Products campaigns "will now be automatically integrated into shopping conversations" with Alexa+. [official] [seat] Prompts on detail pages are a further paid surface on the seller's own page that rivals' campaigns can occupy. Share metrics (§D) are defined on search terms and top-of-search; none of the pages read says whether prompt or Rufus impressions enter them (Open question 9).
+
+2. **GUIDANCE — Amazon's own diagnostics name offer competitiveness.** The 2024-02-01 release note on the campaign diagnostics API covers "featured offer competitiveness" (release-notes/index.md). The Marketing Stream dataset `sponsored-ads-campaign-diagnostics-recommendations` (read 2026-11-09) lists recommendation types (budget, bid, new campaign, budget rules, product-targeting bid, audience bid) and none on competitors. [official] Featured offer and price as preconditions of ad delivery are in `beyond/ad-products-and-eligibility.md`.
+
+3. **GUIDANCE — competitor purchasers in AMC.** The 2025-09-23 release note: the AMC NCS CPG Insights Stream covers "customers who purchase competitor products" (release-notes/index.md). [official] AMC paid data sets are outside a seller's default reach; see `beyond/metrics-and-attribution.md`.
+
+4. **GUIDANCE — share of voice sold outside search.** *unBoxed Toronto 2026 recap and key takeaways* (advertising.amazon.com/library/news/unboxed-toronto-2026, published 2026-03-25, read 2026-11-09) restates RSOV ("Brands can pre-purchase top-of-search placements for branded keywords at a fixed upfront price") and says "The reimagined homepage hero placement now offers flexible share of voice packages and targeting options year-round". [official] The homepage package is not a self-service sponsored-ads product on the pages read; it is listed here only as another place where Amazon sells share of voice by reservation.
+
+### K. Stock-outs and the Featured Offer as competitive moments (gap fill 2026-11-09)
+
+A product's ads depend on its stock and on which seller's offer is featured. Amazon states both rules for one's own ads. The same rules hold for a rival's ads, which makes a rival's stock-out or Featured Offer loss a change in who is bidding.
+
+1. **RULE — Sponsored Products and Display serve only from the SKU that holds the Featured Offer; Sponsored Brands does not need it.** *Understand Featured Offer eligibility* (help GYVDGT5WNNW23LJ5, lastModified 2026-10-16, re-read 2026-11-09): "Failing to hold the Featured Offer directly prevents Sponsored Products and Sponsored Display ads from serving impressions, regardless of campaign status shown in the campaign manager." "Sponsored Products and Display campaigns generate 0 impressions when the advertised ASIN or SKU doesn't hold the Featured Offer, even when the campaign shows "Delivering" status in Amazon Ads Agent." "Sponsored Brands campaigns, including Sponsored Brands video, aren't affected and continue to serve even when another seller holds the Featured Offer, as long as the advertised product is available to buy on Amazon." "When multiple SKUs exist for 1 ASIN, only the SKU that holds the Featured Offer is eligible to serve ads." And: "Featured Offer status can shift to a competitor when their offer becomes more competitive. When status shifts, impressions for the previous holder stop immediately." The eligibility factors it lists include "Product availability : Sufficient stock levels to fulfill orders" and "Price competitiveness : Your total price (item + shipping) compared to other sellers". An Amazon staff reply (Manny_Amazon, Seller Forums thread *Sponsored Products and Buy Box*, re-read 2026-11-09): "Amazon will only display your Sponsored Products ads when your product is the Featured Offer". [official] `beyond/ad-products-and-eligibility.md` §D and `areas/diagnosis.md` carry this gate as a delivery question.
+
+2. **RULE — an out-of-stock product's ads stop.** *How to improve your product detail page for advertising* (advertising.amazon.com/library/guides/improve-your-products-for-advertising, re-read 2026-11-09): "Your products need to be in-stock in order to be the featured offer and be eligible for inclusion in a Sponsored Products campaign. If a product you are advertising goes out of stock, it will be automatically paused within your campaign." Amazon Ads *FAQs* (advertising.amazon.com/resources/faq, re-read 2026-11-09): "If your product isn't the Featured Offer or is out of stock, your ad will not display." [official]
+
+3. **Mechanism — what those two rules mean for a rival's ads.** [seat] The rules are stated for the advertiser, and they bind every advertiser alike. So when a rival's ASIN goes out of stock, the rival's Sponsored Products and Display ads for that ASIN stop serving; and when a reseller's offer takes the Featured Offer on the rival's ASIN, the rival brand's own SKU stops serving SP and SD impressions "immediately", while the reseller's SKU becomes the one that may serve. The rival's Sponsored Brands ads for that product keep serving only "as long as the advertised product is available to buy on Amazon". Each event therefore takes a bidder out of the SP and SD auctions on the queries and pages where the rival advertised that ASIN. No page read says what Amazon shows in the paid slots of a detail page whose product is out of stock or has no Featured Offer, or whether a product target aimed at such a page keeps serving (Open question 18).
+
+4. **Mechanism — the reverse case: a reseller holding one's own Featured Offer.** [seat, from §K.1] When another seller's offer holds the Featured Offer on one's own ASIN, one's own SP and SD ads for that ASIN serve 0 impressions; one's SB ads, including SB video, keep serving and keep sending clicks to a detail page whose default purchase is the other seller's offer. Amazon's page gives the fix for the case where the featured SKU is one's own other SKU ("Add the SKU that holds the Featured Offer. Impressions should resume within 24-48 hours."), and routes the case of another seller to Seller Support ("Advertising Support can't directly restore Featured Offer eligibility"). [official, for the quoted sentences] Removing an unauthorised seller is a Brand Registry and Seller Central matter outside the Ads API (Brand Registry's *Report a Violation*, §Pages read).
+
+5. **GUIDANCE — Amazon's own tools name offer state, not rivals' stock.** The 2024-02-01 release note on campaign diagnostics covers "featured offer competitiveness" (§J.2). No Amazon Ads page or API read reports a rival's stock or Featured Offer state; Amazon's APIs report a rival's offers only through SP-API Product Pricing (§*What cx must read at run time*). [official, seat]
+
+### L. Sponsored Brands video and Store Spotlight in defence and conquest (gap fill 2026-11-09)
+
+1. **RULE — the four Sponsored Brands creative types.** The SB 4.0 model's `CreativeType` enum is `PRODUCT_COLLECTION`, `STORE_SPOTLIGHT`, `VIDEO`, `BRAND_VIDEO` ("The creative type of SB ad"; read 2026-11-09). The *Video ad* API guide (guides/sponsored-brands/ads/video.md, read 2026-11-09): `VIDEO` (`POST /sb/v4/ads/video`) promotes "a single product and link[s] that video to the product detail page"; `BRAND_VIDEO` (`POST /sb/v4/ads/brandVideo`) "promote[s] your brand and link[s] that video to your Store", with "one, two, or three advertised products" under Drive page visits and "none, one, two, or three" under Grow brand impression share; the detail-page video's featured video is "Only horizontal"; vertical video is 9:16, 6–45 seconds. Store Spotlight (`POST /sb/v4/ads/storeSpotlight`) shows the brand logo, a headline and three Store subpages, where a tile click "will direct them to a specified subpage within the Store" and the logo or headline to "your main Store page" (*Store spotlight* API guide). Brand Gallery is an RSOV-only format (§B.6). The enum is not the whole list: the *Ad types overview* guide (guides/sponsored-brands/ads/overview.md, read 2026-11-09) marks "Product collection" "[DEPRECATED]" and names "Collections" (automatic and manual, "3-10 related products"), which the SB 4.0 model creates at `/sb/v4/ads/autoCollection` and `/sb/v4/ads/manualCollection` (US, DE, UK and JP in the features guide). The *Sponsored Brands feature availability by marketplace* guide lists Store Spotlight and both video types in "All marketplaces", and SB product targeting in US, CA, DE, ES, FR, IT, UK, IN and JP. [official] Figures (durations, counts, marketplace lists) are for human readers only.
+
+2. **RULE / GUIDANCE — where each format serves.** *Understand Sponsored Brands* (help GGWFYHL27MFXLHS6, lastModified 2026-10-20): "All ad formats may appear on top of, alongside, or within shopping results, on both desktop and mobile pages". *Campaign goals for Sponsored Brands* (GXXPFXLSVX9VD3U9): under Drive page visits, "All ad formats: Product collection, Store spotlight, and Video", "Placement: Search pages, detail pages, and Amazon homepage"; under Grow brand impression share and Reserve share of voice, "Ad formats with brand Stores only" and "Placement: Top-of-search". *Get started with Sponsored Brands video* (advertising.amazon.com/library/guides/getting-started-with-sponsored-brands-video, datePublished 2021-02-12, dateModified 2026-10-09, read 2026-11-09): "The ads are displayed at the top-of-search and alongside search results or relevant product detail pages on Amazon"; a video linking to the detail page "will appear in shopping results and on product pages"; one linking to the Store "can be featured both above and below the fold of search results". *Sponsored Brands video - Ad specs and guidelines* (advertising.amazon.com/resources/ad-specs/sponsored-brands-video, read 2026-11-09): "Sponsored Brands video ads can also be displayed within detail pages". The *Video ad* API guide is narrower: "Video ads are eligible for mobile and desktop placements in shopping results." [official] So on Amazon's console-facing pages a detail-page-linked video, under Drive page visits, can serve on product pages, which is the placement a product target on a rival's ASIN buys (Open question 20 for the API guide's narrower sentence).
+
+3. **GUIDANCE — Amazon suggests similar brands' names as video keywords, and video has its own auction.** *Get started with Sponsored Brands video*: "Reach shoppers most likely to buy your products by targeting specific keywords or recent product searches. You might try key phrases matching your product description, such as "comfortable walking shoes," or even brand names that are similar to yours." It also names theme targeting, with "keywords related to your brand" to "help build brand loyalty and prevent losing sales". The ad-specs page: "Target shoppers through keywords, categories, and products". *Your complete guide to Sponsored Brands* (datePublished 2021-05-25, dateModified 2026-10-09, re-read 2026-11-09): "Sponsored Brands video has its own ad placement and auction. This means that running a Sponsored Brands video ad campaign won't compete against other self-service ad campaigns you may be running." The SB keyword targeting guide (guides/sponsored-brands/targeting/keyword-targeting.md, read 2026-11-09) has `POST /sb/recommendations/keyword` rank recommendations by `creativeType` ("based on either a static image or a video") and asks `POST /sb/recommendations/bids` callers to pass `adFormat` `productCollection` (for collection and Store Spotlight) or `video` (for video and brand video). [official] [seat] A seller's SBV on a rival's brand term therefore does not bid against its own SP or collection ads on that term, and its bid recommendation is format-specific.
+
+4. **RULE — the creative rules that bound a video on a rival's term or page.** *A complete guide to Sponsored Brands and display ads moderation* (re-read 2026-11-09): "All videos must follow the policies presented in this guide, specifically chapters 1, 2, 3, 4, and 5"; chapter 4 contains "Don't use competitive language": "Ads on Amazon cannot explicitly call out your competitors by name. You may include references to a "leading brand" or other generic phrasing." and "Don't use superiority claims (the best, the biggest, better than ) or mention competitors by name." Chapter 6 (Sponsored Brands): "The keywords or products targeted in your ads must be relevant to the product being promoted." The video chapter: "Video must be relevant to the product or brand advertised, and lead to a contextually relevant landing" page. The ad-specs page: a video campaign "must meet Amazon's creative acceptance policies for sponsored ads". §F.2–F.4 carry the general rules (no disparaging a named competitor; comparative claims need substantiation; no competitive pricing claims). [official] [seat] The targeting may name the rival (its brand term, its ASIN); the video, headline and logo may not.
+
+5. **GUIDANCE (practitioner on an Amazon page) — video for competitor shoppers.** Amazon's *Ad Intel* column *4 common video ad problems and fixes* (advertising.amazon.com/library/expert-advice/four-video-ad-problems-fixes-btr-media, August 26, 2025, by Destaney Wishon, co-founder and CEO of BTR Media, an Amazon Ads Advanced Partner; read 2026-11-09) gives, for "Competitor Shoppers" ("Actively considering a similar product, likely comparison shopping"), the video focus "Product differentiators & brand authority" and the cues "Clear side-by-side comparisons, value callouts, problem-solving features"; and "We've seen Sponsored Brands video drive 32% more conversions than static Sponsored Brands creative in some accounts", footnoted "Advertiser-provided data, US, 2024". [opinion, BTR Media, published by Amazon] (human readers only) [seat] Read beside §L.4, a side-by-side comparison in an Amazon ad may not name the rival and needs substantiation for any superiority claim.
+
+6. **GUIDANCE — Amazon's figure on brand searchers who buy a rival (human readers only).** *Sponsored Brands goal-based campaigns* (advertising.amazon.com/resources/whats-new/sponsored-brands-goal-based-campaigns, October 25, 2023, dateModified 2026-10-09, read 2026-11-09): "From Amazon data of 6,000 advertisers (June 2023), on average, 60% of shoppers who searched for a brand purchased from a competitor brand in the next 30 days from the search. Thus, investing in maintaining and increasing top-of-search impression share will help brands engage with shoppers and build strong brand awareness and consideration." [official] PPC Land's *Sponsored Brands* explainer (2026-10-18) repeats it [opinion]. The page gives no method beyond the sentence, and it is Amazon's own data in the launch of a product it sells.
+
+7. **RULE — how a format's results are read.** v3 has an ad-level Sponsored Brands report (`sbAds`), and the video columns `video5SecondViews` ("The number of impressions where the customer watched the complete video or 5 seconds (whichever is shorter)"), `videoCompleteViews`, `viewabilityRate` and `viewableImpressions` are listed for `sbCampaigns`, `sbAdGroups`, `sbPlacement`, `sbTargeting`, `sbSearchTerm` and `sbAds`; `newToBrandPurchases` and `newToBrandPurchasesPercentage` for SB report types including `sbTargeting` and `sbAds`; `brandStorePageView` for `sbCampaigns` (guides/reporting/v3/columns.md, re-read 2026-11-09). The v3 `adFormat` column is listed for `benchmarks` only. [official] [seat] A per-target read of video on rivals' terms joins `sbTargeting` or `sbAds` rows to the ad's `CreativeType`; the SB bid up-to-300% adjustment applies only under Drive page visits (§E.4).
+
+## What practitioners add
+
+All items here are [opinion]: claims of a named agency, tool vendor or author, stated as theirs. Most of these firms sell software or services that run the tactic they describe; that is noted where it bears on the claim.
+
+### Brand-term bidding: defend, but measure what it adds
+
+1. **Perpetua** (*How to use branded traffic to strategically drive incremental sales on Amazon*, perpetua.io/blog-branded-traffic-strategy-amazon-ads/, undated, no author shown, read 2026-11-09) frames two "truths": "Truth #1: Branded ads can cannibalize organic sales" and "Truth #2: On Amazon, rivals can bid on your brand terms". It says "Branded search is non‑negotiable when your brand queries are materially contested" ("If competitors are appearing above your organic tile or aggressively targeting your PDPs"), names "ASIN mirroring" as a trap to avoid — "when the same product shows up both in the top organic slot and in your top‑of‑search ad", paying "for a click you might have earned for free, without meaningfully increasing your shelf space" — and suggests using branded traffic to "expand coverage rather than duplicate it": complements (refills, accessories, bundles), high-margin and premium lines, launches and overstock rather than the item already ranking first. It gives a "Rule of thumb: If branded spend is >40%, you're likely paying for traffic you'd get organically", says "Branded terms convert at the highest rate and carry the lowest CPCs, which can make advertising cost of sale (ACOS) look artificially strong", and recommends TACoS and brand-halo conversions as the reading instead. The 40% figure is Perpetua's heuristic, with no data shown. (human readers only)
+
+2. **Perpetua** (*Incrementality in Amazon Ads*, perpetua.io/blog-incrementality-amazon-ads-optimization-strategies/, read 2026-11-09): "Branded search traffic… is generally less incremental"; competitor terms are "quite expensive" but have "higher potential for incrementality"; it advises "targeting the weaker competitors with lower ratings and reviews" so that better conversion offsets the high CPC, and favours "'clicky' search terms where click-share is high across the top 3 ASINs" (shoppers not scrolling far), distinguishing "clicky" categories (immediate clicks on top results) from "scrolly" ones.
+
+3. **SalesDuo** (*Amazon Brand Defense PPC: Should You Bid on Your Own Brand Keywords?*, salesduo.com/blog/amazon-brand-defense-ppc/, read 2026-11-09) restates Amazon's guidance and sets the question as "Does paying for branded traffic protect sales that we would otherwise lose?"; "The goal is not maximum branded ROAS. It is efficient protection of incremental branded demand." It warns against "turning off all branded campaigns at once unless you have a controlled testing plan", because seasonality, promotions, inventory, price and competitor moves confound the read, and says "Do not mix branded defense with generic acquisition and competitor conquesting." It names the Sponsored Brands goals (grow brand impression share, RSOV) as defence tools.
+
+4. **Ad Badger** (*How to Segment Branded Keywords in Amazon Ads*, The Badger, published 2019-08-28, modified 2024-05-10, read 2026-11-09) recommends negative-exact of one's own brand name in non-brand campaigns and separate campaigns, with their own budgets, for one's own brand terms and for competitor brand terms. It notes "There's no standard way to know for sure whether a search term is branded or non-branded", so the branded list has to be defined up front, and says "almost 1 in 3 big, brand heavy Amazon sellers are starting to implement this strategy" (the segmentation), with no source for the figure. It also floats a third campaign on competitors' brand terms with a hard daily cap (its example: $50). (human readers only)
+
+5. **m19** (*Does Amazon PPC Sales Cannibalize your Organic Sales?*, m19.com/blog/does-amazon-ppc-sales-really-cannibalize-your-organic-sales, datePublished 2026-10-27 in the page metadata, read 2026-11-09) places cannibalisation where a product "appears twice on top of search results" — paid and organic — and promotes its own "Top of Search Rankings Optimizer" as the remedy. It is a vendor describing its product.
+
+6. **Haus** (*When is branded search worth the investment?*, haus.io/blog/when-is-branded-search-worth-the-investment, read 2026-11-09) is an incrementality-testing vendor reporting a meta-analysis of its clients' geo experiments on **Google** branded search, not Amazon ads. It reports: "82% of brands facing high competition measured significant lift on branded search compared to just 35% of those facing low pressure", with high competition defined as 3 or more competitors totalling over 50% impression share (those brands "surrendered 24% top-position impressions to competitors"); all tests were "conducted on Google text-ad campaigns exclusively targeting brand-name keywords"; the median test ran 14 days with no post-treatment window (27% of tests had one); "For brands selling on Amazon, branded search ads shift sales toward DTC"; and "Don't count on delayed impacts or funnel-building effects from branded search." (human readers only) Its finding that competition decides incrementality matches the Simonov, Nosko & Rao experiment below.
+
+### Share of voice: what the numbers can and cannot say
+
+7. **Trellis** (*Amazon Share of Voice: What to Change When the Number Moves*, gotrellis.com/resources/blog/amazon-share-of-voice, datePublished 2026-10-05 in the page metadata, author not captured, read 2026-11-09): "Amazon publishes no single number called share of voice." It separates three reads: SQP brand share (four stages, a brand view and an ASIN view, "counts combine organic and paid", and only queries the brand already appeared on); STIS, which measures the seller against the field; and top-of-search impression share, whose denominator "is your own eligibility". Its diagnostic: "When they move together, the market moved. When they diverge, you moved." On competitors: identity "has to come from observing the search results page… every tool… inferring from scrapes." It keeps "AI share of voice" — which "counts which brands ChatGPT, Claude and similar assistants recommend for a shopper query" — as "a separate shelf", to be tracked "next to ad share of voice, never inside it". (Trellis sells an AI operator product.)
+
+8. **Intentwise** (*Brand Impression Share report* explainer, intentwise.com/foundation/data-store/amazon-ads/brand-impression-share-report, read 2026-11-09) describes the console dashboard in three sections — Overview, Share over time, Targets — with SP, SB and combined filters, and the report's filters: "brand terms and targets with 1k+ impressions that received less than 20% top-of-search impressions", "with a clickthrough rate of 1.3%+ that received less than 10% of top-of-search impressions", and "with less than 10% top-of-search impressions and a conversion rate of more than 13%". It says "This is a self-service report, and is available to both sellers and vendors." (human readers only; these are Amazon's console filters as Intentwise reports them, not thresholds any page presents as a rule)
+
+9. **Intentwise** (*Are my non-brand Sponsored Ads driving awareness?*, intentwise.com/blog/are-my-non-brand-sponsored-ads-driving-awareness, published 2024-10-15, modified 2026-07-22, read 2026-11-09) argues that non-brand ads do awareness work "covertly", measured in AMC as shoppers who see a non-brand ad and later search the brand; its worked example ("say, 5%") is illustrative, not measured.
+
+10. **Perpetua** (*How to Use the New Amazon Search Term Impression Rank Report*, perpetua.io/blog-amazon-search-term-impression-share-rank-report/, read 2026-11-09) reads impression share as a "share of voice estimate" and notes that the share is account-wide while rows are per campaign and ad group, so impressions must be summed per term; it suggests deriving a relative search-volume estimate from impressions divided by share.
+
+### The detail page as contested ground
+
+11. **AMALYZE** (*Related products and sponsored slots inside the PDP*, amalyze.com/resources/guides/listing/amazon-listing-pdp-related, re-read 2026-11-09): "Your detail page is not just your detail page. Amazon stitches half a dozen recommendation and sponsored carousels into every PDP, and most of them point to other sellers." It maps "Sponsored Products related to this item" as a paid carousel "immediately under the main image block" ("Competitor ads buying targeted impressions on your ASIN. CPCs typically run 1.5–3× higher than SERP Sponsored Products in the same category"), a Sponsored Display slot under Frequently Bought Together, algorithmic carousels it says no one bids into ("Customers who viewed this item also viewed", "Compare with similar items", "Customers who bought this item also bought"), and a brand carousel for Brand-Registered sellers. Its defence: "Run an ad group that targets your own ASIN. Your relevance score is perfect, your CPC is the lowest in the auction, and you occupy the top of the Sponsored Products related carousel", and the same with Sponsored Display. Its audit heuristic: "If your top three direct competitors collectively own more than half the slots across these carousels, you are funding their cost-of-acquisition with your own traffic." It calls the paid carousels "auction-based, second-price". (human readers only; no data shown for the CPC multiple)
+
+11a. **Helium 10** (*Anatomy of a Winning Amazon Product Detail Page*, helium10.com/?p=19922, read 2026-11-09) describes "Sponsored Products Related to this Item" as "where competitors can run product-targeting campaigns to show up on another product's detail page", and places it "above the customer question-and-answer section".
+
+12. **JumpFly** (Reilly McDonnell, *Amazon Brand Protection – You've Got Your Keywords Covered, What About Your ASINs?*, jumpfly.com blog, 2022-03-22, read 2026-11-09): "The product detail pages on Amazon are loaded with ads"; it recommends targeting one's own ASINs with SD, SP and SB; places SD at the top of the page, near the bullets and below the Buy Box, and SP in carousels further down; and says it is "almost impossible to have your ads show up 100% of the time."
+
+13. **AmazOwl** (*Definition: Conquesting Campaigns*, wiki.amazowl.com/sop/amazon-advertising/conquesting-campaigns, read 2026-11-09), writing about "Product Display Ads" (Sponsored Display's former name), defines conquesting as "running adverts on the detail pages of competitive products in order to entice consumers who land there to consider your product instead", next to defensive campaigns on one's own pages, and adds that the advertiser's "page merchandising, images, reviews, pricing etc." have to win the comparison once the shopper clicks.
+
+### Competitive response, bid spikes and the auction dispute
+
+14. **Pacvue** (Tuft & Needle case study, casestudies.com/company/pacvue/case-study/tuft-needle-combats-cpc-spikes-with-pacvue, read 2026-11-09): a spike in branded CPC traced to "a competitor running a 25% off promotion aggressively targeting Tuft & Needle's branded terms"; the response was bid safeguards and dayparting in Pacvue's tool. A vendor case study.
+
+15. **Ad Badger** (*Amazon Says Its Ad Auction Was Second-Price. The FTC Says You Paid Your Own Bid Anyway.*, published 2026-09-10, modified 2026-09-17, read 2026-11-09) summarises both sides of §I and suggests sellers divide average CPC by max bid per keyword ("Ratios clustering at 0.90–1.00 mean you are effectively in a first-price auction"; "0.50–0.75… second-price behavior is still doing work for you") and test a 15% bid cut on 15–20 stable keywords for 14 days. It states that "Sponsored Prompts inventory… became billable on March 25, 2026 and shares your existing campaign bid", with "no separate bid line and no separate budget for prompts"; Amazon's prompts launch page confirms general availability on March 25, 2026 and charging "as part of your CPC bidding and billing parameters" (§J.1). (human readers only)
+
+### Competitor stock-outs and Featured Offer loss (gap fill 2026-11-09)
+
+16. **Perpetua** (Brendan Smith, *How competitor signals help Amazon Advertisers capture market share in real time*, July 28, 2026, perpetua.io/blog-amazon-competitor-signals-for-market-share-growth/, read 2026-11-09) describes its "Market Signals" suite: "With Competitor Signals, Perpetua detects when a competitor goes out of stock and automatically raises bids across your campaigns for the period you specify, then normalizes them once the competitor restocks." On price: "When a competitor becomes less price competitive, your campaigns get more aggressive", while "If a competitor lowers its price in a way that changes the value equation for shoppers, Perpetua can reduce bids to help protect efficiency." On the Featured Offer: "When a competitor loses the Buy Box, they convert less efficiently, which creates an opening. On the other hand, if your own product loses the Buy Box, continuing to spend aggressively can create inefficiency." The page reports no measured result. A vendor describing its own feature.
+
+17. **Jarvio** (Connor Mulholland, *How to Run Amazon Product Targeting Ads (ASIN Targeting)*, 2026-02-28, jarvio.io/blog/amazon-product-targeting-ads, read 2026-11-09): "Frequent stockouts: Track competitors with inconsistent availability. When they're out of stock, their listing page still gets traffic from bookmarks and external links." Its example ASIN list marks one rival "frequently out of stock" with "Target during their stockouts to capture their customers." No data; the page does not say what ads Amazon shows on an out-of-stock page.
+
+18. **Netpeak** (Aliaksandr Vlasenka, *Amazon PPC Strategy: How to Win When Everyone Sells the Same Thing*, 2025-03-31, netpeak.us, read 2026-11-09): "Check their out-of-stock trends – track when they run out of inventory and then pounce."
+
+19. **Eva** (Hai Mag, *Amazon Competitor Keyword Targeting: PPC Strategy & Controls*, published 2024-11-03, modified 2026-11-03, eva.guru/blog/targeting-competitor-keywords-in-amazon-ads/, read 2026-11-09): "A competitor stockout is a window , typically 2 to 6 weeks , where their organic rank is degrading, their listing conversion rate is suppressed, and buyers searching for their brand name are actively choosing between waiting and finding an alternative." It warns of the other side: "An out-of-stock on the ASIN you just spent conquest budget driving traffic to damages the ranking you were trying to build and signals demand instability to Amazon's algorithm", and "A brand with 2 weeks of inventory at current sell-through should not be running an aggressive conquest push." Its decision table lists "Inventory risk | Demand creation could cause a stockout | Cap spend or exclude the SKU". The 2-to-6-week window and the 2-week inventory figure are Eva's, with no data shown (human readers only).
+
+20. **Trellis** (*Stop paying for a lost buy box*, Qore use-case page, gotrellis.com/qore/use-cases/stop-paying-for-a-lost-buy-box, undated, read 2026-11-09): "Losing the Buy Box doesn't stop your ads, it just sends the clicks to a rival", and in its FAQ: "When you lose the Buy Box your Sponsored Products ads keep running, but the clicks land on a listing where a competitor holds the Buy Box" and "Amazon does not pause your ads when you lose the Buy Box". Its product pauses ads on listings where the Buy Box is lost and flags a reprice. For Sponsored Products this contradicts Amazon's help page (§K.1: 0 impressions without the Featured Offer); for Sponsored Brands it matches it.
+
+21. **Canopy Management** (Chuck Kessler, *Amazon Hijackers: How to Remove Unauthorized Sellers*, published 2026-10-09, modified 2026-10-20, canopymanagement.com, read 2026-11-09): "A hijacker hurts most when they take the Featured Offer"; it quotes Amazon as saying "Your offer can't become featured if your item is out of stock" (a Seller Central sentence this seat did not read) and concludes "Every stockout hands the Featured Offer to whoever else is on the listing"; when a hijacker appears it checks "whether a stockout let them in, and whether ad spend is now sending shoppers to someone else's offer".
+
+22. **Pattern UK** (*Losing the Buy Box: The wellness brands losing money to unauthorised resellers*, 2026-07-22, uk.pattern.com, read 2026-11-09): "If a brand doesn't hold its own Buy Box consistently, every pound spent on advertising, every hour spent on content, and every bit of organic ranking work is at least partially benefiting someone else, not the brand that paid for it", and "The more a brand spends on advertising while the Buy Box is fragmented, the more of that spend ends up benefiting competing sellers rather than the brand itself." Pattern sells brand-control services.
+
+23. **A seller, not Amazon staff**, in the Seller Forums (Europe) thread *Run Ads without buy box hijacked listing* (2020, sellercentral-europe.amazon.com, read 2026-11-09): "If you are running Sponsored Brands the ads will continue regardless of who has the buy box, so in that instance anyone selling against that same PDP can benefit from your advertising." This agrees with Amazon's 2026 help page (§K.1).
+
+### Sponsored Brands video and the Store on rival terms (gap fill 2026-11-09)
+
+24. **Skale Strategy** (*Sponsored Brands Video: The Ad Format Most Sellers Ignore*, 2025-10-01, skalestrategy.vercel.app/blog/sponsored-brands-video-ads, read 2026-11-09): "across our portfolio of 100+ accounts, video ads average 40-60% lower cost-per-click than standard Sponsored Products targeting the same keywords"; it structures SBV with "exact match campaigns for high-intent keywords, broad match for discovery, and ASIN targeting to show up on competitor listings", and says "You can own two placements on the same search results page, a Sponsored Products ad and a Sponsored Brands Video ad." It gives no conquest-only figure. It also states SBV ads "link to a single product detail page (not a Store or landing page)", which holds for the `VIDEO` type and not for `BRAND_VIDEO` (§L.1). (human readers only)
+
+25. **ATTN Agency** (*Amazon Sponsored Brands Video: The 2026 DTC Retail Media Goldmine*, 2026-03-12, attnagency.com, read 2026-11-09) gives, as "our 2025 client data", a beauty brand with static SB at "1.2% CTR, $34 CPC, 8% CVR" against video at "2.8% CTR, $31 CPC, 14% CVR", and under SBV targeting, "Supplement brand results": "Generic category keywords: $78 CPC, 9% CVR", "High-intent keywords: $52 CPC, 18% CVR", "Competitor targeting: $45 CPC, 22% CVR". Each figure is one unnamed client, with no dates, volumes or method. [seat] The same page suggests uploading customer lists for lookalike targeting and interest-based targeting, which no Sponsored Brands page read offers; its figures carry that weight. (human readers only)
+
+26. **AMALYZE** glossary entries *Sponsored Brands* and *Sponsored Brands Video* (amalyze.com/resources/glossary/…, undated, read 2026-11-09): competitor brand keywords are "exact match on adjacent competitor brand names. Higher ACOS , conquest play", branded keywords "Defensive, lowest ACOS"; "When a competitor bids on your brand name, the cheapest defence is a Sponsored Brand banner you own outright on your own term"; and "SBV slots in mid-results-page, where shopper intent is high and competitor ad density is lower than at the top", listing among SBV targets "competitor brand names where allowed". No data shown.
+
+27. **Amplisell** (*5 Best Amazon Sponsored Brands Ad Tactics for 2026*, amplisell.com, undated, read 2026-11-09) calls category and competitor ASIN targeting with SB "the primary offensive tactic for growing market share", placing "headline ads on competing product detail pages", and says "When competitors bid on a brand's name, SB headline ads are the most visible counter."
+
+28. **AMALYTIX** (Trutz Fries, *Amazon Sponsored Brands: Formats, Costs, Targeting*, updated 2026-10-31, amalytix.com, read 2026-11-09): "Brand Gallery is only available with Reserve share of voice"; "Amazon only allows a video without any products with the goal "Grow brand impression share"" (which matches the API guide's "none, one, two, or three" under that goal, §L.1); and the Grow brand impression share goal "suits a product launch, Prime Day, or competitors bidding on your brand terms".
+
+29. **Perpetua** (*Amazon Sponsored Brands – Ultimate Guide*, its 2021 guide, perpetua.io/blog-amazon-sponsored-brands-ads-best-practices, read 2026-11-09): the Brand Store "is the only place on Amazon where your competitors can't purchase ad real estate", which is its reason to point Store Spotlight and brand video at the Store; and it advises SB on products "with a history of buy-box ownership", since "Otherwise, your rich content could end up generating sales for your competitor." No Amazon page read states whether a Store page carries other brands' ads (Open question 19).
+
+30. **Measured SBV conquest results.** [seat] The seat found no practitioner or vendor publication that measures Sponsored Brands video on competitor keywords or competitor ASINs separately from video in general, with a stated sample and method. The figures above are single-client or portfolio claims (Skale, ATTN), an Amazon-published partner column with "advertiser-provided data" (BTR Media, §L.5), and Amazon's own case studies of video in general (not opened for this gap). The listed tool vendors' pages that might hold one (Pacvue) returned a Cloudflare challenge or 403 (Unreachable).
+
+### Where practitioners disagree
+
+- **How much brand spend is too much.** Perpetua offers a 40%-of-spend rule of thumb; SalesDuo and Haus say the answer depends on measured competition on the query, not on a share of spend; Amazon's guidance says to defend because competitors "may be bidding". The research below sides with "depends on competition".
+- **What the defence target costs.** AMALYZE says own-ASIN targets win at "the lowest" CPC in the auction and calls the auction second-price; Amazon's own help page says the price "may exceed the runner-up bid" (§I.1), and the FTC alleges winners often pay their own bid (§I.2). No page read measures own-ASIN CPC against competitor CPC on the same page.
+- **Where the paid slots are on a detail page.** AMALYZE puts the Sponsored Products carousel "immediately under the main image block"; Helium 10 puts it "above the customer question-and-answer section"; JumpFly (2022) puts Sponsored Display at the top and Sponsored Products further down. Amazon publishes no layout (Open question 11); the layout varies by device and over time.
+- **Whether own-brand targets perform.** AMALYZE and JumpFly recommend own-ASIN defence targets; Amazon's SB API description says "targeting your own brands usually results in lower performance than targeting competitors' brands" (§E.8a), with no measure stated.
+- **Whether conquest pays.** Perpetua calls competitor terms more incremental but expensive; Amazon suggests conquest where the advertiser's offer is stronger (ratings, reviews, price); the research (Simonov & Hill, on Bing) finds stolen clicks bounce back far more often.
+- **Whether SP keeps running without the Featured Offer.** Trellis's Qore page says Sponsored Products ads "keep running" after a Buy Box loss; Amazon's help page says SP and Display "generate 0 impressions" without it and that only Sponsored Brands keeps serving (§K.1). Perpetua, Canopy and Pattern describe the loss as wasted or leaking spend without separating ad types.
+- **How to react to a rival's stock-out.** Perpetua raises bids "across your campaigns" for the stock-out period; Jarvio and Netpeak aim product targets at the stocked-out rival's page; Eva treats the stock-out as a window of "typically 2 to 6 weeks" and puts one's own inventory cover first. None publishes a measured effect, and Amazon's pages do not say what serves in the paid slots of an out-of-stock page (§K.3).
+- **Where the cheapest defence is.** AMALYZE calls an SB banner on one's own term "the cheapest defence"; Perpetua (2021) calls the Store the only place rivals cannot buy space; Amazon sells a reserved slot (RSOV, §B) and a top-of-search goal (§C). No page read compares their costs on the same terms.
+
+## Published research
+
+All items [research]. For each: what was measured, on what data, and how it transfers to Amazon. Abstracts were read on IDEAS/RePEc, EconPapers, arXiv or the publisher's page on 2026-11-09; where only the abstract was read, the claim is limited to it.
+
+### Own-brand keyword advertising: how incremental is it?
+
+1. **Blake, Nosko & Tadelis (2015), "Consumer Heterogeneity and Paid Search Effectiveness: A Large-Scale Field Experiment", *Econometrica* 83(1), 155–174.** "A series of large‑scale field experiments done at eBay" on paid search. Finding (abstract): "brand keyword ads have no measurable short‑term benefits"; for non-brand keywords, average returns are negative because frequent users, not influenced by ads, account for most spend. Coviello, Gneezy & Goette's abstract summarises the eBay result as "when the company suspended paid search, most of the traffic still ended up on its website". (Which engines were involved, and how many rivals bid on eBay's brand terms, are in the paper's body, not in the abstracts read.) **Transfer:** [seat] general search engines where the brand's own site took the traffic back when ads stopped; read beside the next paper, it looks like the low-competition case, though the abstracts read do not state the competitive state. On Amazon, the organic result for a brand query is the brand's own listing, but rivals' Sponsored Products and Sponsored Brands ads can sit above it, which is the condition the next paper isolates.
+
+2. **Simonov, Nosko & Rao (2018), "Competition and Crowd-Out for Brand Keywords in Sponsored Search", *Marketing Science* 37(2), 200–215.** Experiments on Bing across thousands of brands, varying whether the brand's own ad and competitors' ads were shown. Findings: with no competitors, the brand's own ad adds 1%–4% of clicks, less for larger brands, at a "cost per incremental click" significantly higher than on other keywords; when the focal ad is present, competitors in positions 2–4 "steal" 1%–5%; brands that do not advertise while facing competition lose 18%–42% of clicks; the authors find the "return on investment on defensive advertising to be strongly positive". **Transfer:** the closest published analogue to Amazon brand defence; the magnitudes are for a web engine's text ads and a click outcome, not Amazon purchases, but the mechanism — defence pays in proportion to competitive presence — is the one Amazon's guidance invokes. (human readers only)
+
+3. **Simonov & Hill (2021), "Competitive Advertising on Brand Search: Traffic Stealing and Click Quality", *Marketing Science* 40(5), 923–945.** A large-scale quasi-experimental ad allocation on Bing. Competitors steal 1%–2% of traffic, rising to 6%–15% "if the focal brand's link is exogenously removed from the top paid position"; most stealing is by "a competitor in the top paid link (6%–9% of traffic)". Stolen clicks are lower quality: a "quick back" within 30 seconds happens on 20%–47% of stolen clicks against 7% of the focal brand's clicks. **Transfer:** the conquest side: on brand queries, poached shoppers mostly return. On Amazon the analogue is a rival's ad on a brand query drawing clicks that then fail to convert; Amazon reports conversion for the advertiser, so the conquesting seller sees this directly in its own conversion rate. (human readers only)
+
+4. **Coviello, Gneezy & Goette (2017), "A Large-Scale Field Experiment to Evaluate the Effectiveness of Paid Search Advertising", CESifo WP 6684 / CEPR DP 12333.** The abstract opens on "paid links to branded search terms" and reports a similar experiment at Edmunds.com, "arguably a more representative company": "More than half of the paid traffic is lost when we shut off paid-links search" (EconPapers abstract). **Transfer:** a counter-example to Blake et al. on branded search; the abstract does not state the competitive state of Edmunds' queries. Together with item 1, it shows the answer is site-specific. (human readers only)
+
+### Competition and spillovers among advertisers
+
+5. **Golden & Horton (2021), "The Effects of Search Advertising on Competitors: An Experiment Before a Merger", *Management Science* 67(1), 342–362, doi 10.1287/mnsc.2019.3534.** "Firm Vary" temporarily suspended its Google sponsored search in randomly selected US advertising markets; it later merged with its closest competitor, "Firm Fixed", whose data the authors then used (the firms are unnamed in the abstract). Firm Vary lost customers, but "only 63% as many as a nonexperimental estimate" would say; spillovers to the rival ("Firm Fixed") were "surprisingly small, even in the market for Firm Vary's brand name", the two being "effectively duopsonists". **Transfer:** [seat] in this one experiment a nonexperimental estimate overstated the customer loss from pausing (the abstract does not say what kind of nonexperimental estimate), and a rival did not automatically take what the paused firm gave up. `method/change-measurement.md` does not cite this paper; the directional finding it carries on pauses is the opposite one, from Ju, Zhao & Aral (2025, mobile apps), that a short pause understates a loss that builds through rankings over weeks. The two concern different horizons and mechanisms (a market-level customer count on Google; organic ranking feedback in an app store), so the sign of a naive pause read's error is not settled by either for Amazon.
+
+6. **Lewis & Nguyen (2015), "Display advertising's competitive spillovers to consumer search", *Quantitative Marketing and Economics* 13(2), 93–115.** A natural experiment randomising ad delivery on 500 million Yahoo! homepage visits. In three advertisers' campaigns, display ads raised searches for the advertised brand by 30–45% and for competitors' brands by up to 23%; "the total number of incremental searches for competitors is 2–8 times the increase for advertisers' brands". **Transfer:** upper-funnel ads raise category search that rivals can harvest on their brand and generic terms; on Amazon, the branded-search metrics (§A.7) count only searches for the advertiser's own brand, so this spillover is invisible to the advertiser. (human readers only)
+
+7. **Moshary (2025), "Does Sponsored Search Advertising Augment Organic Search? Evidence from an E-Commerce Platform", *Management Science* 71(11), 9687–9709.** A field experiment blocking all sponsored search for 3% of visitors on an unnamed e-commerce platform: sponsored search "both steals business from organic listings and also reduces the total volume of transactions", yet stays profitable for the platform because ad revenue outweighs lost commissions. **Transfer:** [seat] a platform-level result on one platform: there, sponsored search shifted business away from organic listings and lowered total transactions. For a brand, this is the near-zero-sum frame in which defence and conquest operate. Detail on organic substitution is in `areas/ads-organic-and-total.md`.
+
+8. **Joo, Shi & Abhishek (2024), "Do Sellers Benefit from Sponsored Product Listings? Evidence from an Online Marketplace", *Marketing Science* 43(4), 817–839.** A field study on a mobile marketplace app with experimental and natural variation: consumers prefer top-ranked organic listings to sponsored listings of the same product and position, and become indifferent at lower positions; top organic listings are "perceived as more credible"; a simulation shows a seller "may benefit financially if a lower-ranked organic listing can be replaced with a top-ranked sponsored listing". **Transfer:** [seat] a seller already first organically on its brand query gains little from adding a sponsored copy of the same item there; a seller pushed down by rivals' ads can gain by buying back a top sponsored slot.
+
+9. **Dash et al. (2024), "Sponsored is the New Organic: Implications of Sponsored Results on Quality of Search Results in the Amazon Marketplace", arXiv 2407.19099.** 4,800 simulated searches on Amazon in four countries, 2M organic and 638K sponsored results: items with poor organic rank (beyond 100th) appear as sponsored results "even before the top organic results on the first page", and "in majority of the cases, these top sponsored results are costlier and are of poorer quality than the top organic results" (abstract). **Transfer:** direct Amazon evidence that the results page lets a weaker product buy its way above a stronger one, which is the opening for conquest.
+
+10. **Farronato, Fradkin & MacKay (2025), "Vertical Integration and Consumer Choice: Evidence from a Field Experiment", NBER WP 34135 (2025-08-18).** A field experiment on Amazon that hides Amazon's own brands, and studies their placement in search (NBER page: issued August 2025, revised March 2026): consumers chose similar products, with no extra search effort or cross-platform switching; hiding cost consumers about 5.5% in welfare, and demotion did not raise welfare. **Transfer:** on Amazon, shoppers substitute readily among similar products shown to them, so position on the page is a large part of brand competition. (human readers only)
+
+### Models of buying a rival's keyword
+
+11. **Desai, Shin & Staelin (2014), "The Company That You Keep: When to Buy a Competitor's Keyword", *Marketing Science* 33(4), 485–508.** Game-theory model. A firm "may buy its own brand name only to defend itself"; by buying it, the brand owner "precludes the competitor"; mutual bidding on each other's brands creates "a prisoner's dilemma… the search engine captures the lost profits"; the quality gap between the firms moderates both decisions. **Transfer:** the theoretical frame for a brand-defence standoff: both sellers may be worse off and the platform better off.
+
+12. **Sayedi, Jerath & Srinivasan (2014), "Competitive Poaching in Sponsored Search Advertising and Its Strategic Impact on Traditional Advertising", *Marketing Science* 33(4), 586–608.** Model: "as budget asymmetry increases, the smaller-budget firm poaches more"; the search engine benefits from limiting poaching, which explains relevance handicaps on trademarked keywords. **Transfer:** a smaller rival on a large brand's terms is the predicted pattern. On Amazon, the relevance factors in the auction (§I.1) play the role of the handicap; whether Amazon weighs them specifically against conquest is not stated anywhere read.
+
+13. **Jerath, Ma, Park & Srinivasan (2011), "A 'Position Paradox' in Sponsored Search Auctions", *Marketing Science* 30(4), 612–627.** Model plus Korean search-engine data: a superior firm may bid lower and sit below an inferior firm yet still get more clicks, because informed consumers look for it; the inferior firm has a stronger incentive to be at the top under pay-per-click. **Transfer:** a strong brand need not hold position 1 on a contested generic query; on its own brand query the logic is stronger still, since shoppers came for it. Pairs with Ghose & Yang.
+
+14. **Ghose & Yang (2009), "An Empirical Analysis of Search Engine Advertising: Sponsored Search in Electronic Markets", *Management Science* 55(10), 1605–1622.** A six-month panel of several hundred keywords of a nationwide retailer on Google, modelled with a hierarchical Bayesian simultaneous-equations model (Crossref abstract, re-read 2026-11-09): "conversion rates are highest at the top and decrease with rank", but "profits are often higher at the middle positions than at the top or the bottom ones". **Transfer:** top-of-search share is not automatically the profitable target. This is one of four studies that disagree on how conversion moves with position; `areas/bids-and-placements.md` owns the reconciliation (its P-51): Rutz, Bucklin & Sonnier (2012, a lodging chain, latent instrumental variables) find higher positions raise both click-through and conversion; Agarwal, Hosanagar & Smith (2011, a retailer's field experiment) find "conversion rate increases with position", that is, rises lower down; and Narayanan & Kalyanam (item 14a) show that auction-set positions bias such comparisons. [seat] This file uses Ghose & Yang only for its profit finding, not as the answer on conversion by position.
+
+14a. **Narayanan & Kalyanam (2015), "Position Effects in Search Advertising and their Moderators: A Regression Discontinuity Approach", *Marketing Science* 34(3), 388–407, doi 10.1287/mksc.2014.0893.** Crossref abstract, read 2026-11-09: "Because positions are determined through an auction, there are significant selection issues in measuring position effects. A simple mean comparison of outcomes at two positions is likely to be biased due to these selection issues", and randomising the focal advertiser's bids cannot remove the selection that "competitors' bidding behavior" induces; with a regression discontinuity on a focal advertiser's and its major competitors' daily data, "our RD estimates of these effects are much smaller, and exist only in some of the positions". Moderators: "Position effects are stronger when the advertiser is smaller, and when the consumer has low prior experience with the keyword for the advertiser. They are weaker when the keyword phrase has specific brand or product information, when the ad copy is more specific as in exact matching options, and on weekends compared to weekdays." **Transfer:** [seat] for brand defence, the moderator that matters is the last: on a query that already names the brand, position moved outcomes less in this study, which points the same way as the brand-keyword experiments (items 1–3). The setting is web search text ads, not Amazon product ads.
+
+### Bidding wars and price cycles
+
+15. **Zhang & Feng (2011), "Cyclical Bid Adjustments in Search-Engine Advertising", *Management Science* 57(9), 1703–1719.** Model plus data from two search engines: competing advertisers' bids follow Edgeworth cycles — "price-escalating phases interrupted by price-collapsing phases" — under both first- and second-price auctions. **Transfer:** bid escalation on a contested term followed by collapse is a documented equilibrium pattern, not only a sign of one rival's aggression. Automated bidders on both sides are the setting in which it arises.
+
+16. **Edelman, Ostrovsky & Schwarz (2007), "Internet Advertising and the Generalized Second-Price Auction: Selling Billions of Dollars Worth of Keywords", *American Economic Review* 97(1), 242–259.** GSP is not truthful: truth-telling is not an equilibrium, and bidders shade bids. **Transfer:** the theoretical base under Amazon's "generalized 'second-price' auction" (§I.3); with soft reserves (§I.1) the price rule departs from textbook GSP, so bid-shading logic transfers only loosely.
+
+17. **Brown & MacKay (2023), "Competition in Pricing Algorithms", *American Economic Journal: Microeconomics* 15(2), 109–156.** High-frequency online retail price data and a model: firms with faster, rival-reactive pricing algorithms can sustain higher prices, more dispersion and larger merger effects. **Transfer:** about product prices, not ad bids; relevant because a competitor's price cut on the detail page changes who wins the click and the conversion (Pacvue's case above), and the repricers on Amazon are this kind of algorithm.
+
+### Stock-outs and substitution (gap fill 2026-11-09)
+
+17a. **Conlon & Mortimer (2013), "Demand Estimation Under Incomplete Product Availability", *American Economic Journal: Microeconomics* 5(4), 1–30, doi 10.1257/mic.5.4.1.** Crossref abstract, read 2026-11-09: vending machines with a wireless inventory system tracking availability every four hours; the data "provide valuable variation for identifying substitution patterns when products stock out", and "the corrected model predicts significantly larger impacts of stock-out events on profitability" than a model ignoring availability. **Transfer:** [seat] a stock-out moves buyers to substitutes, and a demand read that ignores availability misjudges the effect; on Amazon the rival's stock-out is a change in the choice set on the rival's queries, and a seller's sales lift in that window is partly substitution that ends when the rival restocks. Physical retail, not search ads.
+
+17b. **Musalem, Olivares, Bradlow, Terwiesch & Corsten (2010), "Structural Estimation of the Effect of Out-of-Stocks", *Management Science* 56(7), 1180–1197, doi 10.1287/mnsc.1100.1170.** Crossref abstract, read 2026-11-09: a structural demand model "that endogenously captures the effect of out-of-stocks on customer choice by simulating a time-varying set of available alternatives", from store-level sales and "partial information on product availability"; it quantifies "the lost sales induced by out-of-stock products" and suggests "price promotion policies that can be used to help mitigate their negative economic impact, which run counter to simple commonly used heuristics". **Transfer:** [seat] the same choice-set logic; neither paper measures advertising during a rival's stock-out, and the seat found no study that does.
+
+### Trademarks in keyword advertising
+
+18. **Bechtold & Tucker (2014), "Trademarks, Triggers, and Online Search", *Journal of Empirical Legal Studies* 11(4), 718–750.** After the CJEU's *Google France* decision, Google relaxed its trademark-keyword policy in continental Europe in September 2010. On 5.38M site visits: little average change; navigational searches became less likely to reach the trademark owner and non-navigational searches more likely; the effect was stronger for lesser-known marks. **Transfer:** opening a brand's name to rivals' bidding mostly hurts small brands on navigational queries — on Amazon, the brand query of a small brand with a loyal searcher base.
+
+19. **Chiou & Tucker (2012), "How Does the Use of Trademarks by Third-Party Sellers Affect Online Search?", *Marketing Science* 31(5), 819–837.** Google's June 2009 change allowing any third-party reseller of a product to use its trademark in ad text, hotel industry: this weakly reduced clicks on the trademark holder's paid ads, a decrease "outweighed by a large increase" in clicks on its unpaid links; when a reseller focuses on the trademark, "the ads become less distinct, and customers are more likely to ignore the advertised offers and buy from the direct channel". **Transfer:** [seat] concerns resellers of the same product using the owner's mark in creative, not rival brands; on Amazon, naming a competitor in creative is forbidden anyway (§F.2). It suggests a mark in someone else's ad text can send shoppers to the owner, but transfers only loosely to Amazon conquest.
+
+### Platform design
+
+20. **Long, Jerath & Sarvary (2022), "Designing an Online Retail Marketplace: Leveraging Information from Sponsored Advertising", *Marketing Science* 41(1), 115–138**, and **Choi & Mela (2019), "Monetizing Online Marketplaces", *Marketing Science* 38(6), 948–972.** Models (Choi & Mela with marketplace data) of a platform trading ad revenue against commissions; Long et al. show bids reveal seller information that the platform can use in organic ranking ("strategic listing"), which intensifies ad competition. **Transfer:** background for why a platform keeps conquest open: ad competition on brand queries is revenue, and in Long et al.'s model it also feeds organic ranking. No claim here about what Amazon's ranking uses.
+
+## Stable enough for a skill
+
+Each item is a fact a skill can carry without a date; figures stay out of the skill and are read at run time. Sources by page name; full URLs in *Pages read*.
+
+1. **Any advertiser may target any brand's name as a keyword and any ASIN as a product target; Amazon lists brand defence ("Protect") and competitor targeting ("Conquest") among its own targeting strategies.** Sources: *A guide to targeting with Sponsored Products*; *What is keyword targeting?*; *Targeting* policy page; *Ad policy change log*.
+2. **Ad creative may not name, disparage or claim superiority over a competitor without substantiation, and may not make competitive price claims; Amazon allows generic phrasing such as "leading brand".** Sources: *Prohibited content and products for Amazon O&O supply*; *Sponsored Brands and display ads moderation guide*; *Claims and substantiation*; *Pricing and Promotions*.
+3. **Reserve share of voice reserves only the advertiser's own exact-match branded keywords from Brand Registry; generic terms, misspellings, competitor brands and trademarks shared by more than two brands are rejected, and keywords, duration and budget cannot change after launch.** Sources: *Branded keyword guidelines and keyword suspension*; *Reserve keywords in a Sponsored Brands campaign*; *Sponsored Brands reserve share of voice* guide; RSOV API guide.
+4. **A Sponsored Brands campaign carries one goal — drive page visits, grow brand impression share, or reserve share of voice — chosen at creation and not changeable.** Sources: *Campaign goals for Sponsored Brands*; SB v4 OpenAPI.
+5. **Search term impression share and rank are account-wide per search term and per ad product (the share of that ad product's impressions on the term across all advertisers) and cannot be broken down by campaign, ad group, target or placement; top-of-search impression share is the campaign's share of its own eligible top-of-search impressions.** Sources: v1 impression-share metric pages; *Downloadable reports in sponsored ads*; *Search term impression share report for Sponsored Products* (what's new); *Top-of-search impression share (top-of-search IS) to measure brand awareness* (what's new).
+6. **No Amazon share metric names the competitors; the Amazon source that names the most-clicked products on a query, with their click and conversion shares, is Brand Analytics' Top Search Terms (top three clicked ASINs). Sponsored Brands brand recommendations (`POST /sb/recommendations/targets/brand`) return suggested brand ids by keyword, with no share.** Sources: v1 impression-share metric pages; *Brand impression share* help; `sellingPartnerSearchTermsReport` schema; SB negative product targeting guide; Trellis (opinion, agreeing on identity).
+7. **Search Query Performance gives the requested ASINs' share of a query's impressions, clicks, cart adds and purchases from the search results page, and requires Brand Registry and the Brand Analytics role; through SP-API the `asin` option is required (space-separated, 200-character limit).** Sources: `sellingPartnerSearchQueryPerformanceReport` schema; SP-API analytics report-type values; sell.amazon.com *Amazon Brand Analytics*. (That the counts combine organic and paid is Trellis's reading [opinion]; the schema says the impressions "originate from the search results page".)
+8. **Negative product and brand targeting keeps Sponsored Products ads off named detail pages, the advertiser's own or a competitor's, and takes effect after a buffer period.** Sources: *Add negative keywords or negative products targeting*; SP negative targeting API guide; SB negative product targeting guide.
+9. **Expanded product targeting is the default on a product target and serves on products Amazon judges closely related; such matches carry a "Product-expanded" prefix in reports.** Source: *Set up product targeting*.
+10. **The matched target report, which shows the detail pages an ad appeared on that drew at least one click, is a Sponsored Display report; for Sponsored Products and Sponsored Brands, the search term report includes only clicked impressions and shows `*` for a detail-page placement with no search keyword.** Sources: *Downloadable reports in sponsored ads*; v1 *Matched target* dimension page; v3 *Search term reports*.
+11. **Branded-search metrics count searches for the advertised brand, including abbreviations and misspellings, attributed to an ad interaction; they exist for Sponsored Brands, Display, TV and DSP, and not for Sponsored Products.** Sources: *Branded searches metric enhancements*; v1 branded-searches metric pages.
+12. **Amazon says the price a winning advertiser pays may exceed the runner-up's bid and never exceeds its own maximum bid, because reserve prices can set it; in Amazon's description, a winner above both reserves pays the soft reserve, and a winner above the hard reserve but below the soft reserve pays its own bid.** Sources: *Understand Amazon Ads auctions*; Amazon's response to the FTC lawsuit. The FTC's allegations about how often the winner pays its own bid are pending.
+13. **In controlled experiments on general search engines, a brand's own ad on its brand query added few clicks when no rival advertised and protected a large share when rivals did.** Sources: Simonov, Nosko & Rao (2018); Blake, Nosko & Tadelis (2015); Coviello, Gneezy & Goette (2017); Simonov & Hill (2021).
+14. **Clicks a rival takes on a brand query bounce back much more often than the brand's own clicks.** Source: Simonov & Hill (2021).
+15. **Competing advertisers' bids on a contested keyword can cycle between escalation and collapse as an equilibrium pattern.** Source: Zhang & Feng (2011).
+16. **Store insights, Brand Metrics and the cross-program benchmarks compare a brand to unnamed peers or category quartiles, never to a named competitor.** Sources: Brand Store insight metrics guide; Brand Metrics overview; Benchmarks report guide.
+17. **Amazon's SB API recommends a seller's own brands for negative targeting, stating that targeting one's own brands "usually results in lower performance than targeting competitors' brands".** Source: SB v4 OpenAPI, `GET /sb/negativeTargets/brands/recommendations`.
+18. **Sponsored Products and Sponsored Brands prompts are on by default in eligible US campaigns, appear in shopping results and on detail pages, and have been charged under the campaign's CPC bidding since general availability on March 25, 2026; prompts with clicks appear in existing SP/SB reporting.** Sources: *Sponsored Products prompts and Sponsored Brands prompts* (what's new); v3 *Prompt Ad Extension reports*.
+19. **Sponsored Products and Display ads for an ASIN serve only from the SKU holding the Featured Offer and only while the product is in stock; when the Featured Offer shifts, the previous holder's impressions stop at once; Sponsored Brands ads, video included, keep serving while the product is available to buy. These rules bind a rival's ads as they bind the seller's.** Sources: *Understand Featured Offer eligibility*; *How to improve your product detail page for advertising*; Amazon Ads *FAQs*; staff reply in *Sponsored Products and Buy Box*.
+20. **When another seller holds the Featured Offer on the seller's own ASIN, the seller's SP and SD ads for it stop and its SB ads continue, sending clicks to a page whose default offer is the other seller's.** Source: *Understand Featured Offer eligibility*.
+21. **Sponsored Brands formats are collections (the older product collection is deprecated; automatic and manual collections replace it), Store Spotlight, video to a detail page (`VIDEO`) and brand video to a Store (`BRAND_VIDEO`), plus RSOV-only Brand Gallery; under Drive page visits every format can serve on search pages, detail pages and the homepage, and under the two top-of-search goals only formats with a Store serve, at top of search.** Sources: SB 4.0 OpenAPI; SB *Ad types overview* and *Video ad* guides; *Campaign goals for Sponsored Brands*; *Understand Sponsored Brands*; *Drive branded search engagement with Sponsored Brands Brand Gallery*.
+22. **Sponsored Brands video runs in its own placement and auction, separate from the advertiser's other self-service campaigns, and can target keywords, categories and products; Amazon's own guide suggests "brand names that are similar to yours" as video keywords.** Sources: *Your complete guide to Sponsored Brands*; *Get started with Sponsored Brands video*; *Sponsored Brands video - Ad specs and guidelines*.
+23. **Every Sponsored Brands and display video must follow the moderation guide's chapters 1–5, which forbid naming competitors and unsubstantiated superiority claims; targeted keywords and products must be relevant to the promoted product.** Source: *A complete guide to Sponsored Brands and display ads moderation*.
+24. **`/v2/stores` was shut off on June 1, 2026 and replaced by `/brand/stores/v1/storePages/list`; `brandEntityId` comes from `POST /brand/stores/v1/stores/list` or SB `GET /brands`.** Sources: *Deprecations*; Brand Home model; SB 3.0 YAML.
+25. **In search advertising, comparing outcomes across auction-set positions is biased by selection; with a regression discontinuity, position effects were smaller than naive estimates, present only at some positions, and weaker on keywords carrying specific brand or product information.** Source: Narayanan & Kalyanam (2015).
+26. **SP-API Product Pricing returns the offers and the featured offer of any ASIN, including other sellers' offers, and up to 20 similar ASINs; Keepa records the history of which seller held the Buy Box and of when an offer was absent.** Sources: Product Pricing v0 and 2022-05-01 models; Keepa `Product.java` and `Stats.java` (vendor data structure).
+
+## What cx must read at run time
+
+All Ads API calls go through the door `cx raw amazon-ads --origin <origin> --method <M> --path <P>` with the profile's scope header, per `method/ads-api-surface.md`; report mechanics per `method/reporting.md`. SP-API and Keepa reads are by name; this seat called none of them.
+
+### Who the brand is, and which terms are "branded"
+
+- **The brand and its ids.** SB `GET /brands` (SB 3.0 YAML; `brandId`, `brandEntityId` "available exclusively for seller advertisers", `brandRegistryName`; referenced from the SB v4 `CreateCampaign.brandEntityId` description); Brand Home `POST /brand/stores/v1/stores/list` (`stores[]` of `brandEntityId`, `storeId`, `storeName`, `storeStatus`) and `POST /brand/stores/v1/storePages/list` (a Store's pages); not `GET /v2/stores`, which the *Deprecations* page shut off on June 1, 2026 even though the Brand Store overview guide still shows it (`method/ads-api-surface.md` owns this); `/adsApi/v1/query/brandStores`; SB brand recommendations `POST /sb/recommendations/targets/brand` for the brand ids used in `asinBrandSameAs` refinements and negatives.
+- **Amazon's own label on keywords.** `POST /adsApi/v1/query/recommendations` with `BRAND_PRODUCT_KEYWORD` (field `keywordTheme`: `Branded`, `Generic`, `Seasonal`, …) and `BRAND_PRODUCT_SEGMENT`.
+- **Reservability as a brand test.** `POST /adsApi/v1/create/keywordReservationValidations/sb` returns `isReservable`, `reservationRejectedReason` and a `keywordReservationValidationId` for each keyword; it is Amazon's own answer to "is this term the seller's brand, someone else's, or generic". It is a create-family call on the v1 surface, its request carries no deal id, and its answer carries a record id; no page read says whether it creates anything billable (Open question 2).
+- **Amazon's own-brand list for negatives.** SB v4 `GET /sb/negativeTargets/brands/recommendations` (on the SB 4.0 model and the JSON "Sponsored Brands 3.0" file, not on the SB 3.0 YAML, whose negative paths are `/sb/negativeKeywords` and `/sb/negativeTargets`; sellers and vendors; US, CA, MX, UK, DE, FR, ES, IT, IN, JP), which Amazon says includes the advertiser's own brands.
+- **The seller's own definition.** Brand names, product lines and misspellings from the listing and the Brand Registry record (SP-API Catalog Items `brand` attribute; Listings Items); the brand list a seller states.
+
+### Competitive state of a query
+
+- **Search term impression share and rank.** Reporting v1 with `searchTerm.value`, `advertiserAccount.id`, `adProduct.value` and `metric.impressionShare`, `metric.impressionShareRank` (no campaign or target dimension); or v3 `spSearchTerm`/`sbSearchTerm` columns "Search Term Impression Share"/"Rank" per `method/reporting.md`. Impressions per term summed across rows.
+- **Top-of-search impression share.** v1 `metric.topOfSearchImpressionShare` at campaign level; v3 `topOfSearchImpressionShare` in `spCampaigns` (grouped by `campaign` or `campaignPlacement`), `sbCampaigns`, `sbTargeting` and `spTargeting` (columns page plus the campaign and targeting report pages; see `areas/bids-and-placements.md` §J).
+- **Keyword recommendations with share.** `POST /sp/targets/keywords/recommendations` (version 4 or later; version 5 is the one Amazon asks integrators to use) fields `searchTermImpressionShare`, `searchTermImpressionRank`; SB keyword recommendations and SB campaign insights with the same pair (SB: "only available if your account has generated impressions for search terms matching the recommended keywords").
+- **Brand Analytics.** SP-API Reports `createReport` with `GET_BRAND_ANALYTICS_SEARCH_TERMS_REPORT` (fields `searchTerm`, `searchFrequencyRank`, `clickedAsin`, `clickShareRank`, `clickShare`, `conversionShare`) and `GET_BRAND_ANALYTICS_SEARCH_QUERY_PERFORMANCE_REPORT` (required `reportOptions.asin`, space-separated, 200-character limit, and `reportPeriod`; fields `searchQuery`, `searchQueryVolume`, `totalQueryImpressionCount`, `asinImpressionShare`, `totalClickCount`, `asinClickShare`, `asinCartAddShare`, `asinPurchaseShare`), then `getReport` and `getReportDocument`. The row for the seller's own brand query is the one that carries the brand-defence read.
+- **Who the competitor products are.** From Top Search Terms `clickedAsin`; SP-API Catalog Items `getCatalogItem` for each ASIN's brand and title; Keepa product fields for price, rank, rating and coupon history of a competitor ASIN, as `cx-keepa`'s `references/product-file.md` defines them (see `areas/launch.md`).
+- **Benchmarks.** v3 `crossProgramBenchmarks` (`groupBy` `brandCategoryBenchmarks`) or v1 benchmark dimensions and P25/P50/P75 metrics.
+
+### Defence and conquest campaigns as they stand
+
+- **Targets on own and competitor ASINs.** SP v1 `POST /adsApi/v1/query/targets` (product targets and their expressions, including expanded); SB `POST /sb/targets/list`; SD `GET /sd/targets` with `asinSameAs`, `asinBrandSameAs`, `views(...)`, `purchases(...)` expressions.
+- **Negatives.** SP v1 `POST /adsApi/v1/query/targets` with `negativeFilter` (per `areas/search-terms-and-negatives.md`), SB `POST /sb/negativeTargets/list` (per `areas/campaign-structure.md`).
+- **Where they served.** v3 `spTargeting` and `spSearchTerm` (clicked impressions only; `*` where a detail-page placement had no search keyword; practitioners read ASIN-valued search terms as product-page matches, which no Amazon page read states); for Sponsored Display only, the matched target report / v3 `matchedTargetAsin` / v1 `matchedTarget.value`; `spPurchasedProduct` for the non-advertised SKUs bought after the click.
+- **Prompts.** v3 `spPromptAdExtension` / `sbPromptAdExtension` (US; 90-day range, 95-day retention), `promptText` and the usual click and sales columns.
+- **Placement.** v3 campaign placement columns (`placementClassification`) for top-of-search, rest-of-search and product pages.
+- **Format of each SB ad.** SB v4 ads list (`POST /sb/v4/ads/list`) for each ad's creative and its `CreativeType` (`PRODUCT_COLLECTION`, `STORE_SPOTLIGHT`, `VIDEO`, `BRAND_VIDEO`; collections ads come from `/sb/v4/ads/autoCollection` and `/sb/v4/ads/manualCollection`, and the enum read does not list a collections value), joined to v3 `sbAds` and `sbTargeting` rows; video columns `video5SecondViews`, `videoCompleteViews`, `viewabilityRate`, `viewableImpressions`; `newToBrandPurchases`, `newToBrandPurchasesPercentage`; `brandStorePageView` in `sbCampaigns`. Format-aware recommendations: `POST /sb/recommendations/keyword` with `creativeType` and `creativeAsins`, and `POST /sb/recommendations/bids` with `adFormat` `productCollection` or `video` (SB 3.0 YAML; SB keyword targeting guide). Store pages for Store Spotlight subpages: Brand Home `storePages/list`, and SB `GET /pageAsins` for the ASINs on a Store page.
+
+### A rival's stock and Featured Offer (gap fill 2026-11-09)
+
+- **Keepa, through `cx-keepa`** (raw `GET /product` with `offers`, `buybox`, `stock` and `stats` where needed, per `cx-keepa`'s `references/raw-routes.md`; read with `references/product-file.md`). Field meanings from Keepa's public API structs (`Product.java`, `Stats.java`, `Offer.java`, `Request.java`, github.com/keepacom/api_backend, read 2026-11-09) [opinion, vendor data structure]:
+  - `buyBoxSellerIdHistory`: "a history of sellerIds that held the Buy Box in the format Keepa time minutes, sellerId"; "If no seller qualified for the Buy Box the sellerId "-1" is used. If it was hold by an unknown seller (a brand new one) the sellerId is "-2""; "Only set if the offers parameter was used".
+  - `stats.buyBoxSellerId`, `stats.buyBoxIsAmazon`, `stats.buyBoxIsUnqualified` ("If there are only sellers with bad offers none qualifies for the buy box"), `stats.buyBoxIsFBA`, `stats.buyBoxIsBackorder`, `stats.buyBoxAvailabilityMessage`, and `stats.buyBoxStats` (per seller id: `percentageWon`, "an approximation of the percentage the seller won the buy box", `avgPrice`, `avgNewOfferCount`, `isFBA`, `lastSeen`); the Buy Box flags are "Only set when the offers parameter was used".
+  - `csv[i]` price histories, where "A price of '-1' means that there was no offer at the given timestamp (e.g. out of stock)"; `stats.outOfStockPercentage30`, `stats.outOfStockPercentage90` and `stats.outOfStockPercentageInInterval` by price slot (`0` "never was out of stock", `-1` "insufficient data or the CsvType is not a price"); `stats.outOfStockCountAmazon30`/`90` ("Number of times in the last 30 days Amazon went out of stock"); `availabilityAmazon` (`-1` no Amazon offer, `0` in stock, `1` preorderable, `2` unknown, `3` backorderable, `4` delayed); `offers[].stockCSV` ("the available stock of this offer as well as, if available, the stock's history") with the `stock` parameter.
+  - `cx-keepa` reads domains US, UK, DE, FR, JP, CA, IT, ES, IN, MX and BR only.
+- **SP-API Product Pricing, through `cx-amazon-sp-api`** (the door's application holds the Pricing role, per its `references/door.md`) [official, models on amzn/selling-partner-api-models, read 2026-11-09]:
+  - `getItemOffers`, `GET /products/pricing/v0/items/{Asin}/offers` ("Returns the lowest priced offers for a single item based on ASIN"; parameters `MarketplaceId`, `ItemCondition`, `CustomerType`), with offers' `SellerId`, `IsBuyBoxWinner` ("When true, the offer is currently in the Buy Box. There can be up to two Buy Box winners at any time per ASIN"), `IsFeaturedMerchant` ("the seller of the item is eligible to win the Buy Box"), `IsFulfilledByAmazon`, `ListingPrice`, `Shipping`, and `Summary` fields `TotalOfferCount`, `NumberOfOffers`, `BuyBoxPrices`, `BuyBoxEligibleOffers`; `getItemOffersBatch` for several ASINs.
+  - `getCompetitiveSummary`, `POST /batches/products/pricing/2022-05-01/items/competitiveSummary`, `includedData` `featuredBuyingOptions` ("the current segmented featured offer information for the item"), `lowestPricedOffers`, `referencePrices` ("such as the Competitive Price Threshold") and `similarItems` ("up to 20 ASINs which are similar to the item"); a featured offer's `offerIdentifier` carries `sellerId`, and "Since this may not belong to the requester, the SKU field is omitted".
+  - Usage plans in the models: 0.5 requests per second for `getItemOffers`, 0.033 for `getCompetitiveSummary` (figures, read from the reply's `x-amzn-RateLimit-Limit` at run time).
+- **One's own ads' state.** The v1 delivery reasons `NOT_IN_BUYBOX` and `OUT_OF_STOCK` on one's own ads, per `areas/diagnosis.md`; Product Pricing reads on one's own SKUs, per `areas/launch.md` §B–C.
+- **What Amazon does not give.** No Ads API read reports a rival's stock or Featured Offer; no read found reports what serves in the paid slots of a rival's out-of-stock page (Open question 18).
+
+### Brand outcomes
+
+- v1 `metric.brandedSearches`, `metric.costPerBrandedSearch`, `metric.brandedSearchRate`, and the new-to-brand family (SB, SD, STV, DSP only).
+- Brand halo columns (`attributionType` in `sbPurchasedProduct`, `asinBrandHalo` in `sdPurchasedProduct`, `…BrandHalo` variants in DSP reports) per `beyond/metrics-and-attribution.md`.
+- Store insights: `POST /stores/{brandEntityId}/insights` (≤100 days), metrics incl. `PEER_DWELL`, `NEW_TO_STORE`, dimension `SOURCE`.
+- Brand Metrics: `POST /insights/brandMetrics/report`, `GET /insights/brandMetrics/report/{reportId}`.
+- Total sales and organic: SP-API Sales and Traffic report (`GET_SALES_AND_TRAFFIC_REPORT`) for the denominator of any brand-spend share, per `areas/ads-organic-and-total.md`.
+
+### Reserve share of voice
+
+- Deals: `POST /adsApi/v1/query/advertisingDeals/sb`, `POST /adsApi/v1/query/advertisingDealTargets/sb`; the campaign's `targetPGDealId` on SB v4 campaigns. Pricing: `POST /adsApi/v1/create/brandedKeywordsPricings/sb` (`keywordsPricing`, `rejectedKeywords`). Current minimums, flight limits and marketplaces from the help pages *Reserve keywords in a Sponsored Brands campaign* and *Branded keyword guidelines*, and the RSOV what's-new page, read at run time because Amazon's own pages disagree.
+
+### Context only a page or the seller gives
+
+- The current state of the FTC case and Amazon's auction description (help GEXGZNBBWPZZK7P8, the two FTC releases, Amazon's response).
+- Current ad-policy pages for creative rules (policies tree G26M9XTK78AMR8TP and its children), and the change log G8X8M375UPCN4UUC.
+- The seller's own brand, trademarks and any agreements with competitors about keywords; the seller's jurisdiction.
+
+## Pages read
+
+All read 2026-11-09 by GET.
+
+| Page name | URL | Type | Date read | What it gave |
+|---|---|---|---|---|
+| Branded keyword guidelines and keyword suspension | https://advertising.amazon.com/help/G2QZJUGUT4RGLJ6N | official, help | 2026-11-09 | RSOV keyword rules, rejection reasons, 14-day mapping, removal |
+| Reserve keywords in a Sponsored Brands campaign | https://advertising.amazon.com/help/G86SD7HK6NHHRB9B | official, help | 2026-11-09 | Frozen fields, 15–92 days, $3,000 minimum, not in Spain, moderation |
+| Campaign goals for Sponsored Brands | https://advertising.amazon.com/help/GXXPFXLSVX9VD3U9 | official, help | 2026-11-09 | Three goals, one per campaign, TOS IS definition, 30–92 days, "strive… 95%" |
+| Brand impression share | https://advertising.amazon.com/help/GNU4AUAUHS76DNDR | official, help | 2026-11-09 | Dashboard scope, brand terms automatic, relative to peers |
+| Understand Amazon Ads auctions | https://advertising.amazon.com/help/GEXGZNBBWPZZK7P8 | official, help | 2026-11-09 | Relevance, reserves, price may exceed runner-up |
+| Add negative keywords or negative products targeting | https://advertising.amazon.com/help/G8W49VU65XQ4T2NS | official, help | 2026-11-09 | Negatives on own/competitor pages, levels, buffers, limits |
+| Set up product targeting | https://advertising.amazon.com/help/GB2JECV9CJK6R6AL | official, help | 2026-11-09 | Expanded default, Product-expanded prefix, refinements |
+| Understand targeting in Display campaigns | https://advertising.amazon.com/help/GPKPWTLW6B8T683N | official, help | 2026-11-09 | Detail-page and audience prospecting |
+| Downloadable reports in sponsored ads | https://advertising.amazon.com/help/GBYSPTSLR337JMLH | official, help | 2026-11-09 | SIS report, CBR, matched target, purchased product |
+| Policies and guidelines | https://advertising.amazon.com/help/G26M9XTK78AMR8TP | official, policy | 2026-11-09 | Policy tree |
+| General requirements | https://advertising.amazon.com/help/G4U9JBZH3664JWX2 | official, policy | 2026-11-09 | Rights to content, Amazon badges |
+| Claims and substantiation | https://advertising.amazon.com/help/G6UU66YLSCUYDULQ | official, policy | 2026-11-09 | Comparative claims, 24-month substantiation |
+| Prohibited content and products for Amazon O&O supply | https://advertising.amazon.com/help/GTEEH3E3J2BDELBA | official, policy | 2026-11-09 | "Disparaging or criticizing a named competitor"; Store titles |
+| Pricing and Promotions | https://advertising.amazon.com/help/GKP5TCJWVN5W7AFU | official, policy | 2026-11-09 | No competitive pricing claims |
+| Prohibited content across all supply | https://advertising.amazon.com/help/GRW23T7LB4MFMBXQ | official, policy | 2026-11-09 | No competitor rule beyond the O&O page |
+| Targeting (policy) | https://advertising.amazon.com/help/GD6SPCK5RJZL73PU | official, policy | 2026-11-09 | No competitor or trademark targeting rule |
+| Ad policy change log | https://advertising.amazon.com/help/G8X8M375UPCN4UUC | official, policy | 2026-11-09 | No 2026 change on competitors/trademarks |
+| How to set up Amazon Sponsored Brands reserve share of voice (guide) | https://advertising.amazon.com/library/guides/sponsored-brands-reserve-share-of-voice | official, library | 2026-11-09 | Beta results, $6K, 14–90 days, 20 marketplaces, FAQ on competitor keywords |
+| Amazon Ads launches reserve share of voice to help brands show up when it matters most (unBoxed 2025 news) | https://advertising.amazon.com/library/news/unboxed-2025-sponsored-brands-reserve-share-of-voice | official, news | 2026-11-09 | 1 in 5 searches branded; ~62% SB TOS IS on branded |
+| Branded searches metric enhancements | https://advertising.amazon.com/resources/whats-new/branded-searches-metric-enhancements | official, what's new | 2026-11-09 | Enhanced matching, not for SP |
+| A guide to targeting with Sponsored Products | https://advertising.amazon.com/library/guides/targeting-with-sponsored-products | official, library | 2026-11-09 | Protect / Conquest strategies |
+| Best practices for your Sponsored Products ads | https://advertising.amazon.com/library/guides/sponsored-products-best-practices | official, library | 2026-11-09 | Target own product pages |
+| What is Keyword Targeting? How It Works, Benefits, Examples | https://advertising.amazon.com/library/guides/keyword-targeting | official, library | 2026-11-09 | Branded keyword definition, other-brand targeting, negative "broad" |
+| Your complete guide to Sponsored Brands | https://advertising.amazon.com/library/guides/sponsored-brands-what-to-know | official, library | 2026-11-09 | Cross-brand shoppers, up-sell |
+| Search term impression share report for Sponsored Products | https://advertising.amazon.com/resources/whats-new/search-term-impression-report-sponsored-products | official, what's new | 2026-11-09 | Account-wide share and rank examples |
+| Introducing the search term impression rank report | https://advertising.amazon.com/resources/whats-new/search-term-rank-report | official, what's new | 2026-11-09 | SB origin of the report, 2020-10-30 |
+| Top-of-search impression share (top-of-search IS) to measure brand awareness | https://advertising.amazon.com/resources/whats-new/top-of-search-impression-share-metric | official, what's new | 2026-11-09 | "eligible to serve on", 90-day lookback |
+| Samsung reserve SOV case study | https://advertising.amazon.com/library/case-studies/sponsored-brands-reserve-sov-samsung | official, case study | 2026-11-09 | 44.9% → 99.8%, footnote "may not be typical" |
+| Bayer reserve SOV case study | https://advertising.amazon.com/library/case-studies/sponsored-brands-reserve-sov-bayer | official, case study | 2026-11-09 | 77% → 97%, Germany, "CPC auction spiral" quote |
+| A complete guide to Sponsored Brands and display ads moderation | https://advertising.amazon.com/library/guides/sponsored-brands-display-ads-moderation | official, library | 2026-11-09 | No competitor names; "leading brand" allowed |
+| Amazon Brand Analytics | https://sell.amazon.com/tools/amazon-brand-analytics | official, sell | 2026-11-09 | Eligibility, Top Search Terms, SQP |
+| Requirements and tips for enrolling a brand in Amazon Brand Registry (sell.amazon.com, reached from the brandservices URL) | https://brandservices.amazon.com/brandregistry | official, sell | 2026-11-09 | Report a Violation exists; no ad-specific protection |
+| Drive top-of-search visibility and engagement by reserving branded share of voice through Sponsored Brands (what's new) | https://advertising.amazon.com/resources/whats-new/sponsored-brands-reserve-share-of-voice | official, what's new | 2026-11-09 | Launch 2025-10-22, modified 2026-10-09; 20 marketplaces incl. EG, not IE |
+| Drive branded search engagement with Sponsored Brands Brand Gallery (what's new) | https://advertising.amazon.com/resources/whats-new/sponsored-brands-brand-gallery | official, what's new | 2026-11-09 | RSOV-only Brand Gallery format, 2026-05-28 |
+| Sponsored Products prompts and Sponsored Brands prompts (what's new) | https://advertising.amazon.com/resources/whats-new/unboxed-2025-sponsored-products-and-sponsored-brands-prompts | official, what's new | 2026-11-09 | GA 2026-03-25, CPC billing, detail-page placement |
+| unBoxed Toronto 2026 recap and key takeaways | https://advertising.amazon.com/library/news/unboxed-toronto-2026 | official, news | 2026-11-09 | Homepage hero share-of-voice packages |
+| Partner Opportunities overview (API) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/recommendations/partner-opportunities/overview.md | official, API | 2026-11-09 | Partner Network membership required |
+| Benchmark brand dimension (v1) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/ads-v1/dimensions/benchmarks/benchmark-brand.md | official, API | 2026-11-09 | benchmarkBrand definition |
+| Matched target dimension (v1) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/ads-v1/dimensions/targeting/matched-target.md | official, API | 2026-11-09 | Sponsored Display only |
+| Search term dimension (v1) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/ads-v1/dimensions/targeting/search-term.md | official, API | 2026-11-09 | SP and SB |
+| Search term reports (v3) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/v3/report-types/search-term.md | official, API | 2026-11-09 | Clicked impressions only; `*` on detail pages without keyword; v3 deprecation 2027-06-30 |
+| Targeting reports (v3) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/v3/report-types/targeting.md | official, API | 2026-11-09 | topOfSearchImpressionShare in SP and SB targeting reports |
+| Reserve share of voice (API guide) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/sponsored-brands/reserve-share-of-voice.md | official, API | 2026-11-09 | Deal lifecycle and operations |
+| Impression share metrics (search) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/ads-v1/metrics/impression-share/search.md | official, API | 2026-11-09 | STIS, rank, TOS IS definitions and compatibility |
+| Branded searches metrics | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/ads-v1/metrics/amazon-retail-conversions/branded-searches.md | official, API | 2026-11-09 | Definitions, eligibility |
+| Benchmarks report (v3) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/v3/report-types/benchmarks.md | official, API | 2026-11-09 | crossProgramBenchmarks fields and limits |
+| Cross-program benchmarks (v1) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/ads-v1/report-types/cross-program-benchmarks.md | official, API | 2026-11-09 | v1 benchmark dimensions |
+| Brand Metrics overview | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/brand-metrics/overview.md | official, API | 2026-11-09 | Peer indices, marketplaces |
+| Brand Metrics how-to | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/brand-metrics/how-to.md | official, API | 2026-11-09 | Endpoints, columns |
+| Share of voice insights (partner opportunities) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/recommendations/partner-opportunities/catalog/share-of-voice-insights.md | official, API | 2026-11-09 | SOV definition, partner-only |
+| SP negative product and brand targeting | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/sponsored-products/negative-targeting/product-brand.md | official, API | 2026-11-09 | Seller campaign-level warning |
+| SB negative product targeting | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/sponsored-brands/targeting/negative-product-targeting.md | official, API | 2026-11-09 | Ad-group level, expression types |
+| SB product targeting | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/sponsored-brands/targeting/product-targeting.md | official, API | 2026-11-09 | Same-as / expanded, up to 300% bid note |
+| SB theme-based targeting | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/sponsored-brands/targeting/theme-based-targeting.md | official, API | 2026-11-09 | KEYWORDS_RELATED_TO_YOUR_BRAND |
+| Brand Store overview | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/brand-store/overview.md | official, API | 2026-11-09 | Store ids |
+| Brand Store insight metrics | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/brand-store/insight-metrics.md | official, API | 2026-11-09 | Metrics incl. PEER_DWELL |
+| Store Spotlight ads | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/sponsored-brands/ads/store-spotlight.md | official, API | 2026-11-09 | Three subpages |
+| SP keyword suggestions | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/sponsored-products/keywords/suggestions.md | official, API | 2026-11-09 | STIS/rank in recommendations v4 |
+| Recommendation types | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/recommendations/recommendations-api/recommendation-types.md | official, API | 2026-11-09 | BRAND_PRODUCT_KEYWORD keywordTheme |
+| Prompt ad extension reports | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/v3/report-types/prompt-ad-extension.md | official, API | 2026-11-09 | Prompts in SP/SB, US only |
+| Release notes | https://d3a0d0y2hgofx6.cloudfront.net/en-us/release-notes/index.md | official, API | 2026-11-09 | Dated launches (benchmarks, RSOV, branded searches, Rufus, TOS IS, prompts columns) |
+| API docs table of contents | https://d3a0d0y2hgofx6.cloudfront.net/en-us/toc2.json | official, API | 2026-11-09 | Page discovery |
+| Sponsored Brands v4 OpenAPI | https://d3a0d0y2hgofx6.cloudfront.net/openapi/en-us/sponsored-brands/4-0/openapi.json | official, API model | 2026-11-09 | Goal enum, IE exclusion, brandEntityId, negative brand recommendations ("targeting your own brands usually results in lower performance") |
+| Amazon Ads API merged OpenAPI (v1) | https://d1y2lf8k3vrkfu.cloudfront.net/openapi/en-us/dest/AmazonAdsAPIALLMerged_prod_3p.json | official, API model | 2026-11-09 | Deal paths and schemas, ThemeMatchType, KPI, BidStrategy enums |
+| Reporting v3 columns | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/v3/columns.md | official, API | 2026-11-09 | Brand halo and branded-search columns |
+| SD audience targeting | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/sponsored-display/audience-targeting.md | official, API | 2026-11-09 | views/purchases expressions, asinBrandSameAs |
+| Sponsored ads campaign recommendations dataset | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/amazon-marketing-stream/datasets/sponsored-ads-campaign-diagnostics-recommendations.md | official, API | 2026-11-09 | Recommendation types; none on competitors |
+| Search Terms report schema | https://raw.githubusercontent.com/amzn/selling-partner-api-models/main/schemas/reports/sellingPartnerSearchTermsReport.json | official, SP-API | 2026-11-09 | Top-3 clicked ASINs fields |
+| Search Query Performance report schema | https://raw.githubusercontent.com/amzn/selling-partner-api-models/main/schemas/reports/sellingPartnerSearchQueryPerformanceReport.json | official, SP-API | 2026-11-09 | Share fields at four stages |
+| Report type values: analytics | https://developer-docs.amazon/sp-api/docs/report-type-values-analytics | official, SP-API | 2026-11-09 | Roles, marketplaces |
+| Amazon's response to the FTC's lawsuit regarding Sponsored Ads | https://www.aboutamazon.com/company-news/amazon-ftc-sponsored-ads-lawsuit-response | official, Amazon response | 2026-11-09 | Hard/soft reserve mechanism (winner below soft reserve pays own bid), 92%, CPC claims |
+| FTC, States Sue Amazon over Secret Ad Surcharge Scheme | https://www.ftc.gov/news-events/news/press-releases/2026/08/ftc-states-sue-amazon-over-secret-ad-surcharge-scheme | regulator | 2026-11-09 | Allegations: soft reserve, ~80% own bid (curl returned 404; read through the WebFetch tool) |
+| FTC, New Details Emerge in Amazon Ads Auction-Rigging Case | https://www.ftc.gov/news-events/news/press-releases/2026/11/new-details-emerge-amazon-ads-auction-rigging-case | regulator | 2026-11-09 | Alleged surcharge levels by season (curl returned 404; read through the WebFetch tool) |
+| FTC, 1-800 Contacts, Inc., In the Matter of | https://www.ftc.gov/legal-library/browse/cases-proceedings/141-0200-1-800-contacts-inc-matter | regulator | 2026-11-09 | Case history, 14 rivals |
+| AIPLA, Second Circuit vacates FTC order in 1-800 Contacts | https://www.aipla.org/detail/news/2021/06/11/u.s.-court-of-appeals-for-the-second-circuit-vacates-final-order-of-the-federal-trade-commission-in-1-800-contacts-case | legal summary | 2026-11-09 | Holding and quotes |
+| FindLaw summary, Multi Time Machine v. Amazon.com | https://caselaw.findlaw.com/summary/opinion/us-9th-circuit/2015/10/21/274719.html | legal summary | 2026-11-09 | Holding on Amazon search results |
+| Simonov, Nosko & Rao 2018 | https://ideas.repec.org/a/inm/ormksc/v37y2018i2p200-215.html | research | 2026-11-09 | Brand-keyword experiment abstract |
+| Simonov & Hill 2021 | https://ideas.repec.org/a/inm/ormksc/v40y2021i5p923-945.html | research | 2026-11-09 | Traffic stealing, click quality |
+| Blake, Nosko & Tadelis 2015 | https://ideas.repec.org/a/wly/emetrp/v83y2015ip155-174.html | research | 2026-11-09 | eBay brand ads |
+| Blake, Nosko & Tadelis (NBER w20171) | https://www.nber.org/papers/w20171 | research | 2026-11-09 | Working-paper version |
+| Coviello, Gneezy & Goette (CESifo 6684) | https://econpapers.repec.org/RePEc:ces:ceswps:_6684 | research | 2026-11-09 | Edmunds experiment abstract |
+| Coviello, Gneezy & Goette (CEPR DP 12333) | https://econpapers.repec.org/RePEc:cpr:ceprdp:12333 | research | 2026-11-09 | Same, CEPR listing |
+| Golden & Horton 2021 | https://ideas.repec.org/a/inm/ormnsc/v67y2021i1p342-362.html | research | 2026-11-09 | Competitor spillovers |
+| Desai, Shin & Staelin 2014 | https://ideas.repec.org/a/inm/ormksc/v33y2014i4p485-508.html | research | 2026-11-09 | When to buy a competitor's keyword |
+| Sayedi, Jerath & Srinivasan 2014 | https://ideas.repec.org/a/inm/ormksc/v33y2014i4p586-608.html | research | 2026-11-09 | Competitive poaching |
+| Zhang & Feng 2011 | https://ideas.repec.org/a/inm/ormnsc/v57y2011i9p1703-1719.html | research | 2026-11-09 | Bid cycles |
+| Ghose & Yang 2009 | https://ideas.repec.org/a/inm/ormnsc/v55y2009i10p1605-1622.html | research | 2026-11-09 | Position and profit |
+| Bechtold & Tucker 2014 | https://ideas.repec.org/a/wly/empleg/v11y2014i4p718-750.html | research | 2026-11-09 | Trademark policy change in Europe |
+| Farronato, Fradkin & MacKay (NBER w34135) | https://www.nber.org/papers/w34135 | research | 2026-11-09 | Amazon brands hidden/demoted |
+| Moshary 2025 | https://ideas.repec.org/a/inm/ormnsc/v71y2025i11p9687-9709.html | research | 2026-11-09 | Ads steal from organic, platform |
+| Joo, Shi & Abhishek 2024 | https://ideas.repec.org/a/inm/ormksc/v43y2024i4p817-839.html | research | 2026-11-09 | Sponsored vs organic by position |
+| Jerath, Ma, Park & Srinivasan 2011 | https://ideas.repec.org/a/inm/ormksc/v30y2011i4p612-627.html | research | 2026-11-09 | Position paradox |
+| Edelman, Ostrovsky & Schwarz 2007 | https://ideas.repec.org/a/aea/aecrev/v97y2007i1p242-259.html | research | 2026-11-09 | GSP properties |
+| Lewis & Nguyen 2015 | https://ideas.repec.org/a/kap/qmktec/v13y2015i2p93-115.html | research | 2026-11-09 | Display spillovers to competitor search |
+| Long, Jerath & Sarvary 2022 | https://ideas.repec.org/a/inm/ormksc/v41y2022i1p115-138.html | research | 2026-11-09 | Marketplace design, strategic listing |
+| Choi & Mela 2019 | https://ideas.repec.org/a/inm/ormksc/v38y2019i6p948-972.html | research | 2026-11-09 | Monetizing marketplaces |
+| Chiou & Tucker 2012 | https://ideas.repec.org/a/inm/ormksc/v31y2012i5p819-837.html | research | 2026-11-09 | Trademarks in third-party ad text |
+| Dash et al. 2024 | https://arxiv.org/abs/2407.19099 | research | 2026-11-09 | Amazon SERP audit |
+| Brown & MacKay 2023 | https://www.aeaweb.org/articles?id=10.1257/mic.20210158 | research | 2026-11-09 | Pricing algorithms |
+| Perpetua, branded traffic strategy | https://perpetua.io/blog-branded-traffic-strategy-amazon-ads/ | opinion, vendor | 2026-11-09 | Two truths, ASIN mirroring as a trap, 40% heuristic |
+| Perpetua, incrementality in Amazon Ads | https://perpetua.io/blog-incrementality-amazon-ads-optimization-strategies/ | opinion, vendor | 2026-11-09 | Branded less incremental; competitor terms |
+| Perpetua, search term impression rank report | https://perpetua.io/blog-amazon-search-term-impression-share-rank-report/ | opinion, vendor | 2026-11-09 | Summing impressions per term |
+| Trellis, Amazon share of voice | https://gotrellis.com/resources/blog/amazon-share-of-voice | opinion, vendor | 2026-11-09 | Three share reads, identity from scrapes |
+| Pacvue, Tuft & Needle case study | https://www.casestudies.com/company/pacvue/case-study/tuft-needle-combats-cpc-spikes-with-pacvue | opinion, vendor | 2026-11-09 | Competitor promo on branded terms |
+| JumpFly, Amazon brand protection: ASINs | https://www.jumpfly.com/blog/amazon-brand-protection-youve-got-your-keywords-covered-what-about-your-asins/ | opinion, agency | 2026-11-09 | Own-ASIN defence across SP/SB/SD |
+| Ad Badger, segment branded keywords | https://www.adbadger.com/blog/how-to-segment-branded-keywords-in-amazon-ads/ | opinion, vendor | 2026-11-09 | Segmentation, no standard branded flag |
+| Ad Badger, second-price vs FTC | https://www.adbadger.com/blog/amazon-says-its-ad-auction-was-second-price-the-ftc-says-you-paid-your-own-bid-anyway/ | opinion, vendor | 2026-11-09 | CPC/bid check; prompts billing claim |
+| m19, PPC cannibalisation | https://www.m19.com/blog/does-amazon-ppc-sales-really-cannibalize-your-organic-sales | opinion, vendor | 2026-11-09 | Double presence at top |
+| Intentwise, Brand impression share report | https://www.intentwise.com/foundation/data-store/amazon-ads/brand-impression-share-report | opinion, vendor | 2026-11-09 | Dashboard sections and preset filters |
+| Intentwise, non-brand ads and awareness | https://www.intentwise.com/blog/are-my-non-brand-sponsored-ads-driving-awareness | opinion, vendor | 2026-11-09 | Unbranded → branded search in AMC |
+| AMALYZE, related products and sponsored slots inside the PDP | https://amalyze.com/resources/guides/listing/amazon-listing-pdp-related | opinion, vendor | 2026-11-09 | Carousel map, own-ASIN defence |
+| Helium 10, Anatomy of a Winning Amazon Product Detail Page | https://www.helium10.com/?p=19922 | opinion, vendor | 2026-11-09 | Page anatomy; places "Sponsored Products Related to this Item" above customer Q&A |
+| Haus, when is branded search worth it | https://www.haus.io/blog/when-is-branded-search-worth-the-investment | opinion, vendor | 2026-11-09 | Google branded-search test meta-analysis |
+| SalesDuo, Amazon brand defense PPC | https://salesduo.com/blog/amazon-brand-defense-ppc/ | opinion, agency | 2026-11-09 | Measure protection, controlled pause |
+| AmazOwl, conquesting campaigns | https://wiki.amazowl.com/sop/amazon-advertising/conquesting-campaigns | opinion, agency | 2026-11-09 | Definition of conquest/defence |
+| Understand Featured Offer eligibility (gap fill) | https://advertising.amazon.com/help/GYVDGT5WNNW23LJ5 | official, help | 2026-11-09 | SP/SD 0 impressions without the Featured Offer; SB unaffected; SKU-specific; "impressions for the previous holder stop immediately"; lastModified 2026-10-16 |
+| Sponsored Products and Buy Box (Seller Forums, Manny_Amazon reply) (gap fill) | https://sellercentral.amazon.com/seller-forums/discussions/t/b55bc259-e982-43bb-a660-58ad413c95ea | official, staff reply | 2026-11-09 | SP ads only when the product is the Featured Offer |
+| How to improve your product detail page for advertising (gap fill) | https://advertising.amazon.com/library/guides/improve-your-products-for-advertising | official, library | 2026-11-09 | Out-of-stock products paused automatically |
+| Amazon Ads FAQs (gap fill) | https://advertising.amazon.com/resources/faq | official | 2026-11-09 | "isn't the Featured Offer or is out of stock, your ad will not display" |
+| Understand Sponsored Brands (gap fill) | https://advertising.amazon.com/help/GGWFYHL27MFXLHS6 | official, help | 2026-11-09 | Formats; all formats on top of, alongside or within results; Store spotlight "at least 4 pages"; lastModified 2026-10-20 |
+| Campaign goals for Sponsored Brands (re-read, gap fill) | https://advertising.amazon.com/help/GXXPFXLSVX9VD3U9 | official, help | 2026-11-09 | Formats and placements by goal; "not available in Spain"; 5–1,000 keywords; 30–92 days |
+| Reserve keywords in a Sponsored Brands campaign (re-read) | https://advertising.amazon.com/help/G86SD7HK6NHHRB9B | official, help | 2026-11-09 | 15–92 days, $3,000, 5–1,000 keywords, not in Spain; lastModified 2026-07-07 |
+| Branded keyword guidelines and keyword suspension (re-read) | https://advertising.amazon.com/help/G2QZJUGUT4RGLJ6N | official, help | 2026-11-09 | 5–1,000 keywords; lastModified 2025-11-05 |
+| How to set up Amazon Sponsored Brands reserve share of voice (re-read) | https://advertising.amazon.com/library/guides/sponsored-brands-reserve-share-of-voice | official, library | 2026-11-09 | $6K US, "e.g., 14–90 days", 20 marketplaces incl. Ireland; published 2025-11-11, modified 2026-10-09 |
+| RSOV what's-new page (re-read) | https://advertising.amazon.com/resources/whats-new/sponsored-brands-reserve-share-of-voice | official, what's new | 2026-11-09 | Marketplace list with Egypt, no Ireland; modified 2026-10-09 |
+| Get started with Sponsored Brands video (gap fill) | https://advertising.amazon.com/library/guides/getting-started-with-sponsored-brands-video | official, library | 2026-11-09 | Placements incl. detail pages; "brand names that are similar to yours"; theme targeting; published 2021-02-12, modified 2026-10-09 |
+| Sponsored Brands video - Ad specs and guidelines (gap fill) | https://advertising.amazon.com/resources/ad-specs/sponsored-brands-video | official, ad specs | 2026-11-09 | Keywords, categories, products; "displayed within detail pages"; acceptance policies |
+| Your complete guide to Sponsored Brands (re-read, gap fill) | https://advertising.amazon.com/library/guides/sponsored-brands-what-to-know | official, library | 2026-11-09 | SBV "has its own ad placement and auction"; Store spotlight three sub-pages |
+| A complete guide to Sponsored Brands and display ads moderation (re-read, gap fill) | https://advertising.amazon.com/library/guides/sponsored-brands-display-ads-moderation | official, library | 2026-11-09 | Videos follow chapters 1–5; chapter 4 competitive language; chapter 6 targeting relevance |
+| Ad Intel: 4 common video ad problems and fixes (BTR Media) (gap fill) | https://advertising.amazon.com/library/expert-advice/four-video-ad-problems-fixes-btr-media | opinion, practitioner on Amazon's site | 2026-11-09 | Competitor shoppers: side-by-side comparisons; SBV +32% conversions, advertiser-provided data |
+| Sponsored Brands goal-based campaigns (what's new) (gap fill) | https://advertising.amazon.com/resources/whats-new/sponsored-brands-goal-based-campaigns | official, what's new | 2026-11-09 | 60% of brand searchers bought a competitor within 30 days (6,000 advertisers, June 2023); 2023-10-25 |
+| SB Ad types overview (API) (gap fill) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/sponsored-brands/ads/overview.md | official, API | 2026-11-09 | Collections, Store spotlight, video; product collection deprecated |
+| SB Video ad (API) (gap fill) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/sponsored-brands/ads/video.md | official, API | 2026-11-09 | VIDEO vs BRAND_VIDEO; product counts by goal; "placements in shopping results" |
+| SB Store spotlight ad (API) (re-read, gap fill) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/sponsored-brands/ads/store-spotlight.md | official, API | 2026-11-09 | Three subpages; click destinations |
+| SB features by marketplace (API) (gap fill) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/sponsored-brands/features.md | official, API | 2026-11-09 | Format and targeting marketplaces; negative paths on `sponsored-brands/3-0/openapi` |
+| SB keyword targeting (API) (gap fill) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/sponsored-brands/targeting/keyword-targeting.md | official, API | 2026-11-09 | `creativeType` in keyword recommendations; `adFormat` in bid recommendations |
+| Sponsored Brands 3.0 (YAML model) (gap fill) | https://d3a0d0y2hgofx6.cloudfront.net/openapi/en-us/sponsored-brands/3-0/openapi.yaml | official, API model | 2026-11-09 | `/brands`, `/sb/negativeKeywords`, `/sb/negativeTargets`, `/pageAsins`; no brand-negative recommendations path |
+| Sponsored Brands 3.0 (JSON file) (gap fill) | https://d1y2lf8k3vrkfu.cloudfront.net/openapi/en-us/dest/SponsoredBrands_prod_3p.json | official, API model | 2026-11-09 | 50 paths; `/sb/negativeTargets/brands/recommendations`; no `/sb/negativeKeywords` |
+| Brand Home model (gap fill) | https://d1y2lf8k3vrkfu.cloudfront.net/openapi/en-us/dest/BrandHome_prod_3p.json | official, API model | 2026-11-09 | `stores/list` (`StoreInfo.brandEntityId`), `storePages/list` |
+| Deprecations (gap fill) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/release-notes/deprecations.md | official, API | 2026-11-09 | V2 Stores: deprecated January 2026, shut off June 1, 2026 |
+| Campaign reports (v3) (gap fill) | https://d3a0d0y2hgofx6.cloudfront.net/en-us/guides/reporting/v3/report-types/campaign.md | official, API | 2026-11-09 | topOfSearchImpressionShare under SP campaign/campaignPlacement and SB campaign |
+| Product Pricing v0 model (gap fill) | https://raw.githubusercontent.com/amzn/selling-partner-api-models/main/models/product-pricing-api-model/productPricingV0.json | official, SP-API model | 2026-11-09 | `getItemOffers`, `IsBuyBoxWinner`, `IsFeaturedMerchant`, `BuyBoxEligibleOffers` |
+| Product Pricing 2022-05-01 model (gap fill) | https://raw.githubusercontent.com/amzn/selling-partner-api-models/main/models/product-pricing-api-model/productPricing_2022-05-01.json | official, SP-API model | 2026-11-09 | `getCompetitiveSummary`, `featuredBuyingOptions`, `similarItems`, `offerIdentifier` |
+| Keepa API structs: Product.java, Stats.java, Offer.java, Request.java (gap fill) | https://github.com/keepacom/api_backend/tree/master/src/main/java/com/keepa/api/backend/structs | opinion, vendor data structure | 2026-11-09 | `buyBoxSellerIdHistory`, Buy Box stats, out-of-stock percentages and counts, `availabilityAmazon`, `stockCSV`, request parameters |
+| Perpetua, competitor signals (gap fill) | https://perpetua.io/blog-amazon-competitor-signals-for-market-share-growth/ | opinion, vendor | 2026-11-09 | Bids up on competitor stock-out; Buy Box loss as an opening; 2026-07-28 |
+| Jarvio, product targeting ads (gap fill) | https://jarvio.io/blog/amazon-product-targeting-ads | opinion, vendor | 2026-11-09 | Target rivals during stock-outs; 2026-02-28 |
+| Netpeak, PPC when everyone sells the same thing (gap fill) | https://netpeak.us/blog/amazon-ppc-strategy-how-to-win-when-everyone-sells-the-same-thing/ | opinion, agency | 2026-11-09 | "track when they run out of inventory and then pounce" |
+| Eva, competitor keyword targeting (gap fill) | https://eva.guru/blog/targeting-competitor-keywords-in-amazon-ads/ | opinion, vendor | 2026-11-09 | Stock-out window "2 to 6 weeks"; own inventory risk |
+| Trellis, Stop paying for a lost buy box (gap fill) | https://gotrellis.com/qore/use-cases/stop-paying-for-a-lost-buy-box | opinion, vendor | 2026-11-09 | Says SP keeps running after Buy Box loss (contradicts Amazon) |
+| Canopy Management, Amazon hijackers (gap fill) | https://canopymanagement.com/amazon-hijackers-unauthorized-sellers/ | opinion, agency | 2026-11-09 | Stock-outs hand the Featured Offer to others; ad spend to someone else's offer |
+| Pattern UK, losing the Buy Box (gap fill) | https://uk.pattern.com/blog/losing-the-buy-box-the-wellness-brands-losing-money-to-unauthorised-resellers | opinion, agency | 2026-11-09 | Fragmented Buy Box sends ad value to resellers |
+| Seller Forums (Europe), Run Ads without buy box hijacked listing (gap fill) | https://sellercentral-europe.amazon.com/seller-forums/discussions/t/c12ad27b21f719c3e4273a09ffc4968b | opinion, seller posts | 2026-11-09 | SB continues regardless of the Buy Box holder |
+| Skale Strategy, Sponsored Brands Video (gap fill) | https://skalestrategy.vercel.app/blog/sponsored-brands-video-ads | opinion, agency | 2026-11-09 | 40–60% lower CPC than SP on same keywords; ASIN targeting on competitor listings |
+| ATTN Agency, Sponsored Brands Video (gap fill) | https://www.attnagency.com/blog/amazon-sponsored-brands-video | opinion, agency | 2026-11-09 | Single-client CTR/CVR figures incl. competitor targeting |
+| AMALYZE glossary: Sponsored Brands; Sponsored Brands Video (gap fill) | https://amalyze.com/resources/glossary/sponsored-brands | opinion, vendor | 2026-11-09 | Competitor brand keywords as conquest; SBV mid-page |
+| Amplisell, Sponsored Brands ad tactics (gap fill) | https://amplisell.com/blog-posts/best-amazon-sponsored-brands-ad-tactics | opinion, agency | 2026-11-09 | Competitor ASIN targeting as offensive tactic |
+| AMALYTIX, Sponsored Brands formats (gap fill) | https://www.amalytix.com/en/knowledge/marketing/amazon-sponsored-brands/ | opinion, vendor | 2026-11-09 | Brand Gallery RSOV-only; video without products under BIS |
+| Perpetua, Sponsored Brands ultimate guide (2021) (gap fill) | https://perpetua.io/blog-amazon-sponsored-brands-ads-best-practices | opinion, vendor | 2026-11-09 | Store as the place rivals cannot buy ads |
+| PPC Land, Sponsored Brands explainer (gap fill) | https://ppc.land/sponsored-brands/ | opinion, trade press | 2026-11-09 | Repeats Amazon's 60% figure; format history |
+| Ghose & Yang 2009; Narayanan & Kalyanam 2015; Agarwal, Hosanagar & Smith 2011; Rutz, Bucklin & Sonnier 2012; Golden & Horton 2021 (Crossref, contradiction re-check) | https://api.crossref.org/works/10.1287/mnsc.1090.1054 (and the other DOIs in *Published research*) | research | 2026-11-09 | Verbatim abstracts |
+| Conlon & Mortimer 2013 (Crossref) (gap fill) | https://api.crossref.org/works/10.1257/mic.5.4.1 | research | 2026-11-09 | Stock-out substitution, vending data |
+| Musalem et al. 2010 (Crossref) (gap fill) | https://api.crossref.org/works/10.1287/mnsc.1100.1170 | research | 2026-11-09 | Structural out-of-stock model |
+
+Sibling files re-read (local, not public pages, not counted): `areas/keyword-and-product-targeting.md`, `areas/bids-and-placements.md`, the cx-amazon-ads skill and door reference, and the listing-playbook `paid-keywords.md` research.
+
+**Unreachable**
+
+- *Google France v Louis Vuitton*, CJEU C-236/08 on EUR-Lex (https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:62008CJ0236 and the HTML variant) — empty response to GET; cited only through Bechtold & Tucker.
+- *The data behind incrementality on Amazon* (Search Engine Land, https://searchengineland.com/the-data-behind-incrementality-on-amazon-330294, and its MarTech copy) — Cloudflare challenge, then 403 (re-tried by the verifier with curl and WebFetch on 2026-11-09: 403).
+- VoxEU column on branded search ads (https://cepr.org/voxeu/columns/experiment-estimate-effectiveness-branded-search-ads; voxeu.org copy) — 403 (re-tried by the verifier: 403).
+- The "additional terms that apply to Reserve Share of Voice campaigns" linked from the RSOV help pages — not opened by either seat.
+- Wilson Sonsini note on the 1-800 Contacts decision (wsgr.com) — 403.
+- IP Update note on the same decision (ipupdate.com) — empty response.
+- *Multi Time Machine v. Amazon.com* on Justia (law.justia.com, July and October 2015 opinions) — 403; read through FindLaw's summary instead.
+- Second Circuit 1-800 Contacts opinion PDF on govinfo — binary PDF; the local PDF tools fail on a missing library, so the text was not extracted; read through AIPLA's summary instead.
+- Seller Central help page for Search Query Performance (G8J4CB5ZBF3NX7TP) and other Seller Central help pages tried (G10471, G200418100, G202059240) — a plain GET serves only a JavaScript shell; the content call is a POST this seat may not make.
+- FTC press-release search page — empty result to GET.
+- Amazon Ads blog *display ads boost growth for search advertisers* — the URL served a generic library landing page, not the article.
+- Web search budget was exhausted mid-session; later reads were of known URLs only, so some practitioner pages (Skai, SmartScout, SellerSprite) were not reached at all.
+- Gap fill 2026-11-09: Pacvue *Amazon advertising and brand protection: what they have in common* and *Weeks of cover* blog posts (pacvue.com) — Cloudflare challenge to curl, 403 to WebFetch; a search summary attributes to Pacvue that SB, DSP and offsite spend "will continue, regardless of the Buy Box", which this seat could not read on the page.
+- Gap fill 2026-11-09: LandingCube *Amazon Ads Competitor Product Targeting: A Practical Conquest Strategy for 2026* (landingcube.com/?p=511304) and Seller Metrics *Defending Your Digital Shelf Against AI Bidding Bots* (sellermetrics.app) — bot-verification pages to curl and WebFetch.
+- Gap fill 2026-11-09: Search Engine Land *Why your Amazon Ads aren't delivering: 6 critical issues to fix* — Cloudflare challenge.
+- Gap fill 2026-11-09: Seller Central help pages on the Featured Offer (the "Your offer can't become featured if your item is out of stock" sentence Canopy quotes) — not attempted; a plain GET serves a JavaScript shell and the content call is the forbidden POST.
+
+## Open questions
+
+1. **Reserve share of voice flight length and minimum.** Amazon's own pages give 15–92 days, 30–92 days and "e.g., 14–90 days", and $3,000 versus $6K minimum (US). Which governs today is not stated; a run-time read of the help page and the pricing response is the only check.
+2. **Is `keywordReservationValidations` side-effect-free?** The request carries no deal id and the answer carries validity, a reason and a `keywordReservationValidationId`; the API guide calls it a check ("call this API to check whether they can be reserved") and marks it optional, but it sits in the v1 `create` family and returns a record id, and no page says whether it creates anything billable.
+3. **RSOV marketplaces.** The library guide lists 20 including IE; the what's-new launch page (modified 2026-10-09) lists 20 with EG instead of IE; the SB v4 model says goal-based campaigns are not supported in IE; two help pages (G86SD7HK6NHHRB9B and GXXPFXLSVX9VD3U9) exclude Spain (absent from both lists). The library guide and the what's-new page carry the same modification date, 2026-10-09, so neither is the later word.
+4. **An API for the Brand Impression Share dashboard.** None found in the TOC, the merged v1 model or the release notes. Whether the dashboard's "relative to peers" figure and "lost top-of-search sales" are reachable outside the console is open.
+5. **Partner share-of-voice insights for a seller's own app.** The insight is for Partner Network members; whether a seller-authorised tool like cx qualifies is not stated.
+6. **Seller campaign-level negative products on manual SP campaigns.** The API guide's warning (vendors only) conflicts with the 2025-05-22 release note (every campaign, US) and the SP 3.0 model (AUTO only), as recorded in `areas/search-terms-and-negatives.md`. Which is current is open.
+7. **Negative broad.** A library guide mentions negative broad; the help page lists only negative phrase and exact.
+8. **Is the soft reserve still applied, and how?** Amazon's help page (lastModified 2026-04-27) still describes reserves that set price; the FTC case is pending. Whether anything changed after 2026-08-31 is not stated on the pages read.
+9. **Prompts, Rufus and Alexa+ in share metrics.** No page read says whether prompt or Rufus impressions count in STIS, TOS IS or brand impression share. (The billing date is settled: Amazon's prompts launch page gives general availability and CPC charging from 2026-03-25; the prompt report page says "Prompts with clicks will show in your existing Sponsored Products or Sponsored Brands reporting", without saying which metrics.)
+10. **Expanded targets on own-ASIN defence.** Expanded product targeting is default; whether an own-ASIN target with expansion on spends materially on competitors' pages is measurable only from the seller's own reports ("Product-expanded" rows in the targeting column; the search term report for SP; the matched target report for SD only — see question 17). No public measurement found.
+11. **The detail-page ad layout.** Amazon publishes no map of where Sponsored Products and Sponsored Display ads sit on a detail page; practitioner maps disagree and change by device and over time.
+12. **Brand query definition.** Amazon's branded-search metrics use "enhanced brand matching" including misspellings, RSOV rejects misspellings, and reports carry no branded flag on a search term. A single Amazon definition that a seller's reports can be split by does not exist on the pages read.
+13. **Trademark law by jurisdiction.** US case law read here (MTM v. Amazon; 1-800 Contacts) and the EU's Google France decision (unreached) cover keyword use; other jurisdictions where cx sellers operate (UK, JP, CN-based sellers on US marketplaces) were not researched.
+14. **Incrementality of brand defence on Amazon specifically.** The controlled studies are on Bing, Google, eBay and Edmunds; the Amazon-specific evidence read is Amazon's own before/after case studies of attributed sales and vendor meta-analyses. No independent Amazon holdout study of brand-term ads was reached (the Search Engine Land article on Amazon incrementality was blocked).
+15. **Which ad product takes the peer-brands theme.** `KEYWORDS_RELATED_TO_PEER_BRANDS_PRODUCT_CATEGORY` is in the v1 model's shared `ThemeMatchType`, `MatchType` and `TargetMatchTypeFilter` enums; no guide read says whether Sponsored Products, Sponsored Brands or both accept it.
+16. **Own-brand targets in Sponsored Brands.** Amazon's SB API says targeting one's own brands "usually results in lower performance than targeting competitors' brands" without naming the metric; whether that refers to ROAS, conversion, incrementality or something else is not stated.
+17. **Detail-page matches in Sponsored Products reports.** Amazon's matched-target report and dimension are Sponsored Display only; for Sponsored Products the pages read say only that detail-page placements without a search keyword show `*`. Which SP field identifies the detail page a product target served on is not stated on the pages read.
+18. **Paid slots on a rival's out-of-stock or Featured-Offer-less page.** Amazon states that an out-of-stock or non-featured product's own SP and SD ads stop. No page read says whether other advertisers' product targets on that page keep serving, or what fills the paid slots there; Jarvio's claim that such pages "still get traffic from bookmarks and external links" is unmeasured. A seller's own `spTargeting` rows for a product target on that ASIN, before and during the rival's stock-out (with Keepa's out-of-stock history for the dates), are the only evidence in reach.
+19. **Do Brand Store pages carry other brands' ads?** Perpetua (2021) says the Store is "the only place on Amazon where your competitors can't purchase ad real estate"; no Amazon page read states it either way.
+20. **Sponsored Brands video on detail pages.** The console-facing pages (*Get started with Sponsored Brands video*, the ad-specs page, *Campaign goals* under Drive page visits) place SB video on product detail pages; the API's *Video ad* guide says video ads "are eligible for mobile and desktop placements in shopping results" without mentioning detail pages. Whether a `BRAND_VIDEO` or a `VIDEO` ad with a product target on a rival's ASIN serves on that rival's page is answered by the seller's own `sbTargeting` and placement rows, not by a page read.
+21. **Measured SBV conquest.** No published, method-stated measurement of Sponsored Brands video on competitor keywords or ASINs, separate from video in general, was found (§*What practitioners add* item 30).
+22. **What Trellis's "SP keeps running" claim rests on.** Trellis's Qore page says SP ads keep running after a Buy Box loss; Amazon's Featured Offer page (lastModified 2026-10-16) says 0 impressions. Whether Trellis means a partial Buy Box share (the SKU holding the Featured Offer part of the day, which Amazon's staff reply's "at the time the customer views the ad" wording, recorded in `areas/diagnosis.md`, allows) is not stated on its page.
+
+## Positions
+
+**P-brand-defense-and-competition-1.** Amazon lets any advertiser target any brand's name as a keyword and any ASIN as a product target, and names brand defence ("Protect") and competitor targeting ("Conquest") among its targeting strategies. Sources: [official] *A guide to targeting with Sponsored Products*; *What is keyword targeting?*; *Targeting* policy page (no competitor rule).
+
+**P-brand-defense-and-competition-2.** Amazon's guidance is to consider bidding on one's own branded keywords "as a brand defense strategy since your competitors may be bidding on your branded keywords, too". Source: [official] *A guide to targeting with Sponsored Products*.
+
+**P-brand-defense-and-competition-3.** Ad creative on Amazon may not name or disparage a competitor, may not make unsubstantiated superiority claims or competitive price claims, and may refer to a "leading brand" generically; health-product comparative claims need dated substantiation. Sources: [official] *Prohibited content and products for Amazon O&O supply*; *Sponsored Brands and display ads moderation guide*; *Claims and substantiation*; *Pricing and Promotions*.
+
+**P-brand-defense-and-competition-4.** Reserve share of voice reserves the Sponsored Brands top-of-search slot at a fixed price for the advertiser's own exact-match branded keywords only; competitor brands, generic terms, misspellings and trademarks shared by more than two brands are rejected. Sources: [official] *Branded keyword guidelines and keyword suspension*; *Sponsored Brands reserve share of voice* guide.
+
+**P-brand-defense-and-competition-5.** Once a reserve share of voice campaign starts, its keywords, duration and budget cannot change; a proposed deal cannot be modified, and it can be deleted only until twelve hours before its start, which archives its campaigns. Sources: [official] *Reserve keywords in a Sponsored Brands campaign*; RSOV API guide.
+
+**P-brand-defense-and-competition-6.** Amazon's pages state different reserve share of voice limits — flights of 15–92, 30–92 or "e.g., 14–90" days, and a US minimum of $3,000 or $6K — while all three help pages agree on 5 to 1,000 reserved keywords (figures as of 2026-11-09). Sources: [official] *Reserve keywords in a Sponsored Brands campaign*; *Campaign goals for Sponsored Brands*; *Branded keyword guidelines and keyword suspension*; *Sponsored Brands reserve share of voice* guide.
+
+**P-brand-defense-and-competition-7.** A Sponsored Brands campaign has one goal — drive page visits (CPC, all placements), grow brand impression share (vCPM, top-of-search) or reserve share of voice (fixed price) — set at creation and not changeable. Sources: [official] *Campaign goals for Sponsored Brands*; SB v4 OpenAPI.
+
+**P-brand-defense-and-competition-8.** Amazon reports its reserve share of voice beta (15+ advertisers, May–August 2025) raised Sponsored Brands top-of-search impression share on reserved keywords from 62.7% to 99.3%; these are Amazon's own before/after figures, not holdout tests. Sources: [official] *Sponsored Brands reserve share of voice* guide; Samsung and Bayer case studies.
+
+**P-brand-defense-and-competition-9.** Search term impression share is the account's share of one ad product's ad impressions on a search term across all advertisers (Amazon's example: 20% means "you won 20% of all Sponsored Products ad impressions for that search term on that date"), and its rank is the account's position among advertisers; both are account-wide and cannot be split by campaign, ad group, target or placement. Sources: [official] v1 impression-share metric pages; *Downloadable reports in sponsored ads*; *Search term impression share report for Sponsored Products* (what's new).
+
+**P-brand-defense-and-competition-10.** Top-of-search impression share is a campaign's top-of-search impressions divided by the top-of-search impressions it was eligible for, so its denominator is the campaign's own eligibility, not the whole field. Sources: [official] v1 impression-share metric pages; *Top-of-search impression share metric* (what's new); [opinion] Trellis.
+
+**P-brand-defense-and-competition-11.** The console's Brand impression share dashboard shows top-of-search impression share on brand terms Amazon identifies automatically, across the brand's SP and SB campaigns and accounts, relative to peers; no API for it was found. Sources: [official] *Brand impression share* help; [opinion] Intentwise.
+
+**P-brand-defense-and-competition-12.** None of Amazon's share metrics names competitors; the Brand Analytics Top Search Terms report names the three most-clicked ASINs on a query, with click and conversion shares, and Sponsored Brands brand recommendations return suggested brand ids by keyword without shares. Sources: [official] `sellingPartnerSearchTermsReport` schema; v1 impression-share metric pages; SB negative product targeting guide; [opinion] Trellis.
+
+**P-brand-defense-and-competition-13.** Search Query Performance reports the requested ASINs' share of a query's impressions, clicks, cart adds and purchases on the search results page, by week, month or quarter, for Brand Registry sellers with the Brand Analytics role; through SP-API the report requires an `asin` list of at most 200 characters. Sources: [official] `sellingPartnerSearchQueryPerformanceReport` schema; SP-API report type values (analytics); sell.amazon.com *Amazon Brand Analytics*. Trellis reads its counts as organic and paid combined [opinion, Trellis].
+
+**P-brand-defense-and-competition-14.** Amazon recommends targeting one's own product pages and bidding higher on them to keep one's range visible there. Source: [official] *Best practices for your Sponsored Products ads*.
+
+**P-brand-defense-and-competition-15.** Expanded product targeting is the default on a product target and also serves on products Amazon judges closely related; such matches are labelled "Product-expanded" in reports. Source: [official] *Set up product targeting*.
+
+**P-brand-defense-and-competition-16.** Negative product and brand targeting keeps Sponsored Products ads off named detail pages, the advertiser's own or a competitor's, with a buffer of 72 hours for negative keywords and 96 hours for negative products (figures as of 2026-11-09). Sources: [official] *Add negative keywords or negative products targeting*; SP negative targeting API guide.
+
+**P-brand-defense-and-competition-17.** Amazon's API guide says campaign-level negative product targeting on manual Sponsored Products campaigns is for vendors only, while a 2025 release note widens it; the two pages conflict. Sources: [official] SP negative product and brand targeting guide; release notes (as recorded in `areas/search-terms-and-negatives.md`).
+
+**P-brand-defense-and-competition-18.** The matched target report, listing the product detail pages an ad appeared on that drew at least one click, exists for Sponsored Display only; Sponsored Products and Sponsored Brands search term reports include only clicked impressions and show `*` where a detail-page placement had no search keyword. Sources: [official] *Downloadable reports in sponsored ads*; v1 *Matched target* dimension page; v3 *Search term reports*.
+
+**P-brand-defense-and-competition-19.** Sponsored Display can target audiences who viewed or bought products of a given brand (`asinBrandSameAs`) and detail pages of chosen products, own or other. Sources: [official] SD audience targeting guide; *Understand targeting in Display campaigns*.
+
+**P-brand-defense-and-competition-20.** The Amazon Ads v1 theme `KEYWORDS_RELATED_TO_PEER_BRANDS_PRODUCT_CATEGORY` is defined as search terms shoppers use for "products from other brands in the category of your advertised products", with "The peer brands are selected automatically"; the pages read do not say which ad product accepts it. Source: [official] Amazon Ads API merged OpenAPI, `ThemeMatchType`.
+
+**P-brand-defense-and-competition-21.** Branded-search metrics count searches for the advertised brand, including abbreviations and misspellings, attributed to an ad interaction, for Sponsored Brands, Display, TV and DSP, and are not available for Sponsored Products. Sources: [official] *Branded searches metric enhancements*; v1 branded-searches metric pages.
+
+**P-brand-defense-and-competition-22.** Amazon reports no branded/non-branded flag on a search term; its keyword recommendations carry a `keywordTheme` label such as `Branded` or `Generic`, and practitioners define the branded list themselves. Sources: [official] Recommendation types guide; [opinion] Ad Badger, *How to Segment Branded Keywords*.
+
+**P-brand-defense-and-competition-23.** Amazon states that the price a winning advertiser pays can exceed the runner-up's bid because reserve prices can set it, and never exceeds the advertiser's maximum bid; in its own description, a winner whose bid clears both reserves pays the soft reserve, and a winner whose bid clears the hard reserve but not the soft reserve still wins and pays its own bid. Sources: [official] *Understand Amazon Ads auctions*; Amazon's response to the FTC lawsuit.
+
+**P-brand-defense-and-competition-24.** The FTC and 22 states allege (August 2026, pending) that Amazon's soft reserve made Sponsored Products winners pay their own bid close to 80% of the time; Amazon answers that about 92% of selected Sponsored Products ads in 2024 were not the highest bid and that no advertiser pays more than its bid (figures as of 2026-11-09, human readers only). Sources: [official, Amazon response] aboutamazon.com; [official, regulator — allegations] FTC press releases of 2026-08-31 and 2026-11-05.
+
+**P-brand-defense-and-competition-25.** In Bing experiments across thousands of brands, a brand's own ad on its brand query added 1%–4% of clicks when no competitor advertised, while brands that did not advertise against competitors lost 18%–42%. Source: [research] Simonov, Nosko & Rao (2018).
+
+**P-brand-defense-and-competition-26.** In eBay's field experiments, brand-keyword ads had "no measurable short-term benefits" (when eBay suspended paid search, most of the traffic still reached its site); at Edmunds.com, shutting off paid links to branded search lost more than half the paid traffic. Sources: [research] Blake, Nosko & Tadelis (2015); Coviello, Gneezy & Goette (2017).
+
+**P-brand-defense-and-competition-27.** On Bing, competitors' ads on a brand query took 1%–2% of its traffic, rising to 6%–15% when the brand's own link was removed from the top paid position, and 20%–47% of stolen clicks returned to Bing within 30 seconds against 7% of the brand's own. Source: [research] Simonov & Hill (2021).
+
+**P-brand-defense-and-competition-28.** In an experiment before a merger, a firm that paused its Google search ads in random US markets lost only 63% as many customers as a non-experimental estimate implied, and spillovers to its closest rival were "surprisingly small", even on the paused firm's brand name. Source: [research] Golden & Horton (2021).
+
+**P-brand-defense-and-competition-29.** Game-theory models find that mutual bidding on each other's brand names can be a prisoner's dilemma in which the platform gains, and that the smaller-budget firm poaches more as budget asymmetry grows. Sources: [research] Desai, Shin & Staelin (2014); Sayedi, Jerath & Srinivasan (2014).
+
+**P-brand-defense-and-competition-30.** Competing advertisers' bids on a keyword can follow Edgeworth cycles of escalation and collapse under both first- and second-price auctions. Source: [research] Zhang & Feng (2011).
+
+**P-brand-defense-and-competition-31.** An audit of 4,800 simulated Amazon searches in four countries found products with organic rank beyond 100 shown as sponsored results above the top organic results, and in most cases top sponsored results costlier and "of poorer quality" than top organic ones. Source: [research] Dash et al. (2024).
+
+**P-brand-defense-and-competition-32.** When Google opened trademarks to rivals' keyword bidding in continental Europe in 2010, navigational searches became less likely to reach the trademark owner, more so for lesser-known marks. Source: [research] Bechtold & Tucker (2014).
+
+**P-brand-defense-and-competition-33.** The Ninth Circuit held in 2015 that Amazon showing other brands' watches for a search of "mtm special ops" was not likely to confuse shoppers, because Amazon labels each result with its name, manufacturer and photo; the Second Circuit in 2021 vacated the FTC's order against 1-800 Contacts' settlement agreements requiring rivals to refrain from search advertising on its trademarks and to use them as negative keywords. Sources: [legal] FindLaw summary of *Multi Time Machine v. Amazon.com*; AIPLA summary of *1-800 Contacts v. FTC*; [official, regulator] FTC matter page.
+
+**P-brand-defense-and-competition-34.** Perpetua's rule of thumb is that branded spend above 40% of total ad spend likely pays for traffic the brand would get organically, and that branded terms make ACOS look "artificially strong". Source: [opinion] Perpetua, *How to use branded traffic*.
+
+**P-brand-defense-and-competition-35.** Haus reports, from Google branded-search tests, that 82% of brands facing high competition on their brand terms measured significant lift against 35% facing low competition. Source: [opinion] Haus, *When is branded search worth the investment?*.
+
+**P-brand-defense-and-competition-36.** Store insights carry a peer dwell-time comparison, Brand Metrics indices compare a brand to category peers, and the cross-program benchmarks give peers' 25th, 50th and 75th percentiles; none names a peer. Sources: [official] Brand Store insight metrics guide; Brand Metrics overview; Benchmarks report guide.
+
+**P-brand-defense-and-competition-37.** Amazon's Sponsored Brands API recommends the advertiser's own brands for negative targeting, stating that "targeting your own brands usually results in lower performance than targeting competitors' brands"; the statement names no metric and shows no data. Source: [official] SB v4 OpenAPI, `GET /sb/negativeTargets/brands/recommendations`.
+
+**P-brand-defense-and-competition-38.** Sponsored Products and Sponsored Brands prompts are enrolled automatically, can appear in shopping results and on product detail pages, have been charged under the campaign's CPC bidding in the US since general availability on March 25, 2026, and can be paused one by one in the console. Sources: [official] *Sponsored Products prompts and Sponsored Brands prompts* (what's new); v3 *Prompt Ad Extension reports*; [opinion] Ad Badger (agreeing on the date).
+
+**P-brand-defense-and-competition-39.** Amazon's RSOV pages list different marketplaces: the library guide includes Ireland, the what's-new launch page (modified 2026-10-09) includes Egypt instead, the SB v4 model excludes goal-based campaigns in Ireland, and two help pages exclude Spain (lists as of 2026-11-09). Sources: [official] *How to set up Amazon Sponsored Brands reserve share of voice*; RSOV what's-new page; SB v4 OpenAPI; *Reserve keywords in a Sponsored Brands campaign*.
+
+**P-brand-defense-and-competition-40.** Sponsored Products and Sponsored Display campaigns "generate 0 impressions when the advertised ASIN or SKU doesn't hold the Featured Offer", only the SKU holding the Featured Offer serves ads for an ASIN, and "When status shifts, impressions for the previous holder stop immediately"; Sponsored Brands, including video, keeps serving while the product "is available to buy on Amazon". Sources: [official] *Understand Featured Offer eligibility* (help GYVDGT5WNNW23LJ5); staff reply in *Sponsored Products and Buy Box*.
+
+**P-brand-defense-and-competition-41.** Amazon states that an advertised product that goes out of stock "will be automatically paused within your campaign" and that an ad does not display if the product "isn't the Featured Offer or is out of stock"; these rules apply to every advertiser, so a rival's stock-out or Featured Offer loss stops that rival's SP and SD ads for the ASIN, and no Amazon page read says what then serves in the paid slots of the rival's page. Sources: [official] *How to improve your product detail page for advertising*; Amazon Ads *FAQs*; *Understand Featured Offer eligibility*.
+
+**P-brand-defense-and-competition-42.** When another seller holds the Featured Offer on a brand's own ASIN, the brand's SP and SD ads for that ASIN serve no impressions while its Sponsored Brands ads keep serving and send shoppers to a page whose default offer is the other seller's; Amazon routes that case to Seller Support, saying "Advertising Support can't directly restore Featured Offer eligibility". Sources: [official] *Understand Featured Offer eligibility*; [opinion] Canopy Management (Chuck Kessler), Pattern UK and a seller in the Seller Forums (Europe), on spend reaching the other seller's offer.
+
+**P-brand-defense-and-competition-43.** Keepa's data structure records the Buy Box holder as a time series (`buyBoxSellerIdHistory`, "-1" when no seller qualified, "-2" for an unknown new seller, set only when offers are requested), per-seller Buy Box win shares (`stats.buyBoxStats.percentageWon`), price histories in which "-1" means "no offer at the given timestamp (e.g. out of stock)", and 30-day, 90-day and interval out-of-stock percentages. Source: [opinion, vendor data structure] Keepa `Product.java`, `Stats.java`, `Offer.java` (github.com/keepacom/api_backend), read through `cx-keepa`'s product-file reference.
+
+**P-brand-defense-and-competition-44.** SP-API's `getItemOffers` returns an ASIN's lowest-priced offers with each `SellerId` and `IsBuyBoxWinner`, and `getCompetitiveSummary` returns its featured buying options (whose `offerIdentifier` "may not belong to the requester"), lowest-priced offers, reference prices and "up to 20 ASINs which are similar to the item". Source: [official] Product Pricing v0 and 2022-05-01 models (amzn/selling-partner-api-models).
+
+**P-brand-defense-and-competition-45.** Perpetua's Competitor Signals "detects when a competitor goes out of stock and automatically raises bids across your campaigns for the period you specify, then normalizes them once the competitor restocks", and Perpetua says a competitor that loses the Buy Box will "convert less efficiently, which creates an opening"; Jarvio and Netpeak advise product-targeting a rival's page during its stock-outs; Eva calls a rival's stock-out a window of "typically 2 to 6 weeks" and warns that conquest-driven sales can stock out the conquering ASIN. None of the four publishes a measured effect. Sources: [opinion] Perpetua (Brendan Smith, 2026-07-28); Jarvio (Connor Mulholland, 2026-02-28); Netpeak (Aliaksandr Vlasenka, 2025-03-31); Eva (Hai Mag, modified 2026-11-03).
+
+**P-brand-defense-and-competition-46.** Trellis's Qore page says Sponsored Products ads "keep running" after a Buy Box loss, while Amazon's Featured Offer page says SP and SD serve 0 impressions without the Featured Offer and only Sponsored Brands keeps serving. Sources: [official] *Understand Featured Offer eligibility*; [opinion] Trellis, *Stop paying for a lost buy box*.
+
+**P-brand-defense-and-competition-47.** Sponsored Brands formats are collections (product collection is marked deprecated; automatic and manual collections replace it), Store Spotlight (a Store with a home page and at least three subpages, each with a unique product), video linking one product to its detail page (`VIDEO`) and brand video linking to the Store (`BRAND_VIDEO`, one to three products under Drive page visits, none to three under Grow brand impression share); Brand Gallery serves only with reserve share of voice. Sources: [official] SB *Ad types overview*, *Video ad* and *Store spotlight* API guides; SB 4.0 OpenAPI `CreativeType`; *Understand Sponsored Brands*; *Your complete guide to Sponsored Brands*; *Drive branded search engagement with Sponsored Brands Brand Gallery*.
+
+**P-brand-defense-and-competition-48.** Under the Drive page visits goal every Sponsored Brands format can serve on search pages, detail pages and the Amazon homepage, and under Grow brand impression share and reserve share of voice only formats with a Store serve, at top of search; Amazon's video guide and ad-specs page place SB video on product detail pages as well as in search results, while the API video guide names only shopping-results placements. Sources: [official] *Campaign goals for Sponsored Brands*; *Get started with Sponsored Brands video*; *Sponsored Brands video - Ad specs and guidelines*; SB *Video ad* API guide.
+
+**P-brand-defense-and-competition-49.** Amazon says Sponsored Brands video "has its own ad placement and auction" and "won't compete against other self-service ad campaigns you may be running", targets "keywords, categories, and products", and suggests as video keywords "even brand names that are similar to yours"; SB bid recommendations take an `adFormat` of `productCollection` or `video`. Sources: [official] *Your complete guide to Sponsored Brands*; *Get started with Sponsored Brands video*; *Sponsored Brands video - Ad specs and guidelines*; SB keyword targeting API guide.
+
+**P-brand-defense-and-competition-50.** Every Sponsored Brands and display video must follow chapters 1 to 5 of Amazon's moderation guide, whose chapter 4 says "Ads on Amazon cannot explicitly call out your competitors by name" while allowing "a "leading brand" or other generic phrasing" and forbids unsubstantiated superiority claims; Sponsored Brands targets "must be relevant to the product being promoted". Source: [official] *A complete guide to Sponsored Brands and display ads moderation*.
+
+**P-brand-defense-and-competition-51.** Amazon's 2023 goal-based campaigns launch page states that, "From Amazon data of 6,000 advertisers (June 2023), on average, 60% of shoppers who searched for a brand purchased from a competitor brand in the next 30 days from the search" (figure for human readers only). Sources: [official] *Sponsored Brands goal-based campaigns* (what's new, 2023-10-25); [opinion] PPC Land, repeating it.
+
+**P-brand-defense-and-competition-52.** Published figures on Sponsored Brands video against competitors are practitioner claims without a stated method: Skale Strategy reports video CPCs "40-60% lower" than Sponsored Products on the same keywords across "100+ accounts", ATTN Agency reports one unnamed supplement client's competitor targeting at "$45 CPC, 22% CVR", and BTR Media, on Amazon's Ad Intel page, reports video driving "32% more conversions than static Sponsored Brands creative in some accounts" from "Advertiser-provided data"; the seat found no measurement of SB video on competitor terms or ASINs alone (figures for human readers only). Sources: [opinion] Skale Strategy (2025-10-01); ATTN Agency (2026-03-12); Destaney Wishon, BTR Media, *Ad Intel* (2025-08-26).
+
+**P-brand-defense-and-competition-53.** Perpetua's 2021 Sponsored Brands guide calls the Brand Store "the only place on Amazon where your competitors can't purchase ad real estate"; no Amazon page read states whether Store pages carry other brands' ads. Source: [opinion] Perpetua, *Amazon Sponsored Brands – Ultimate Guide*.
+
+**P-brand-defense-and-competition-54.** Amazon's Deprecations page lists `/v2/stores` as deprecated in January 2026 and shut off on June 1, 2026, replaced by `/brand/stores/v1/storePages/list`; `brandEntityId` is returned by `POST /brand/stores/v1/stores/list` and by Sponsored Brands `GET /brands` (for sellers), and the API's Brand Store overview guide still shows `GET /v2/stores` (as of 2026-11-09). Sources: [official] *Deprecations*; Brand Home model; Sponsored Brands 3.0 YAML; Brand Store overview guide.
+
+**P-brand-defense-and-competition-55.** `/sb/negativeKeywords` and `/sb/negativeTargets` are on the YAML model "Amazon Ads API - Sponsored Brands" version 3.0, and `/sb/negativeTargets/brands/recommendations` is on the SB 4.0 model and on the JSON file titled "Sponsored Brands 3.0", neither of which carries `/sb/negativeKeywords`. Source: [official] the three Sponsored Brands models, parsed 2026-11-09.
+
+**P-brand-defense-and-competition-56.** The v3 columns page lists `topOfSearchImpressionShare` for `spCampaigns`, `sbCampaigns` and `sbTargeting`, and the v3 targeting and campaign report pages list it for the Sponsored Products and Sponsored Brands targeting reports and for SP and SB campaign reports. Source: [official] Reporting v3 *columns*, *Targeting reports* and *Campaign reports* pages.
+
+**P-brand-defense-and-competition-57.** Narayanan and Kalyanam find that comparing outcomes across auction-set ad positions is biased by selection, that regression-discontinuity position effects are "much smaller" and "exist only in some of the positions", and that they are weaker "when the keyword phrase has specific brand or product information"; the four studies of conversion by position (Ghose & Yang; Rutz, Bucklin & Sonnier; Agarwal, Hosanagar & Smith; Narayanan & Kalyanam) disagree, as `areas/bids-and-placements.md` records. Sources: [research] Narayanan & Kalyanam, *Marketing Science* 2015; Ghose & Yang, *Management Science* 2009; Rutz, Bucklin & Sonnier, *JMR* 2012; Agarwal, Hosanagar & Smith, *JMR* 2011.
+
+**P-brand-defense-and-competition-58.** In stock-out studies outside advertising, a model that accounts for availability predicts significantly larger stock-out effects on profitability than one that ignores it, and both studies model a stock-out as a change in the set of alternatives buyers choose among. Sources: [research] Conlon & Mortimer, *AEJ: Microeconomics* 2013; Musalem et al., *Management Science* 2010.
+
+### Seat's positions (enter no skill)
+
+- [seat] Three Amazon metrics called "impression share" use three denominators (all advertisers on a term; the campaign's own eligibility; all ads on a reserved keyword's slot), so a reader comparing them across reports compares different things.
+- [seat] The research converges on "brand defence pays in proportion to competitive presence on the brand query"; on Amazon the presence can be observed from search term impression share and rank on the brand term and from Top Search Terms' clicked ASINs, so the competitive state is a run-time read, not a constant.
+- [seat] A defence campaign on own ASINs with expanded targeting on may be partly a conquest campaign; the "Product-expanded" rows of the seller's reports (and, for Sponsored Display, the matched-target report) are the evidence.
+- [seat, practitioner knowledge not on an Amazon page read] ASIN-valued rows in the Sponsored Products search term report are read by practitioners as the detail pages a product target matched.
+- [seat] Under either side's account of the auction, a bid raise on a contested term can raise the price paid without a rival moving, which weakens "outbid the competitor by a cent" as a cost model on Amazon.
+- [seat] The conquest-side research (Simonov & Hill's bounce-backs; loosely, Chiou & Tucker on resellers' trademark ads) and Amazon's conquest guidance (target where your offer is stronger) point the same way: conquest's return is set by the comparison the shopper makes after the click.
+- [seat] A rival's stock-out or Featured Offer loss changes the auction twice over: the rival's SP and SD ads for that ASIN leave it (Amazon's rule), and the rival's shoppers face a smaller choice set (the stock-out studies). The first is a fact of Amazon's serving rules; the second, and any sales lift a seller sees in that window, ends when the rival restocks or regains the offer, so a lift read across the window is not a lift the seller keeps.
+- [seat] A brand whose own Featured Offer is held by a reseller still pays for Sponsored Brands clicks to that page; its own SP and SD stop by themselves. The facts that bear on the SB spend are which ad types serve and whose offer the click lands on.
+- [seat] In Sponsored Brands video on a rival's term or page, the targeting may name the rival (its brand keyword, its ASIN) and the creative may not; "leading brand" phrasing and substantiated claims are what the moderation guide leaves to a comparison video.
+- [seat] Store Spotlight and brand video send the click to the seller's own Store; a detail-page video sends it to a detail page, which AMALYZE describes as carrying rivals' sponsored carousels (practitioners item 11). Whether a Store page carries no other brands' ads is a practitioner claim, not an Amazon one (Open question 19).
+
+## Verification
+
+Verified 2026-11-09 by an adversarial verifier seat.
+
+**Method.** Every page in *Pages read* was re-opened on 2026-11-09 by plain GET, none sampled. Amazon Ads help pages: the `window.__scV2Content` JSON in the page HTML was decoded and its `lastModified` checked against the file's. The help page *Downloadable reports in sponsored ads* was parsed table-row by table-row to read which ad product each report covers. Amazon Ads API guides were read through the CloudFront markdown host (`d3a0d0y2hgofx6.cloudfront.net/en-us/<path>.md`). The merged v1 OpenAPI (`d1y2lf8k3vrkfu.cloudfront.net/.../AmazonAdsAPIALLMerged_prod_3p.json`) and the SB v4 OpenAPI were parsed for the exact enums, schemas and paths cited. The full release-notes index was split by entry and searched by date and keyword. SP-API schemas came from `raw.githubusercontent.com/amzn/selling-partner-api-models` and the public report-type page. Library, case-study and what's-new pages, aboutamazon.com, sell.amazon.com and practitioner pages were read as HTML text, with `datePublished`/`dateModified` taken from page metadata where present. The two FTC press releases returned HTTP 404 to curl and were read with the WebFetch tool; the FTC matter page, the AIPLA and FindLaw summaries, the NBER and AEA pages were also read with WebFetch. Research claims were checked against the abstract on IDEAS/RePEc, EconPapers, arXiv, NBER or AEA, and anything beyond the abstract was cut back or marked. To find 2025–2026 announcements that might overturn a claim, the verifier searched the web and Amazon's what's-new pages for news on RSOV, prompts, soft reserves and brand share. That turned up the RSOV and Brand Gallery what's-new pages, the prompts GA page and the unBoxed Toronto 2026 recap. No change to soft reserves after 2026-08-31 was found. Each position was checked against the wording rule (a fact, a rule, a mechanism, a measured effect or a named claim; never a when-to-ask list, a fixed sequence, a threshold presented as a rule, a ceiling, or how cx talks).
+
+**Corrections (was → now).**
+1. §E.10, stable item 10, P-18, run-time "Where they served": the matched target report was presented as the read of a product target's pages in general → it is a **Sponsored Display-only** report (help page table; v1 `matchedTarget.value` lists only SD). For SP/SB, the v3 search term report has clicked impressions only, with `*` for detail-page placements that have no keyword. Reading ASIN-valued rows as product-page matches is moved to the seat's list as practitioner knowledge. Open questions 10 and 17 were updated or added.
+2. §H.4 and run-time: `asinBrandHalo` was quoted as "whether a purchase is attributed to a promoted product or brand-halo effect" → that is the description of `attributionType` (`sbPurchasedProduct`). `asinBrandHalo` is "a product that was purchased after an ad click but has a different SKU/ASIN than what was advertised" (`sdPurchasedProduct`).
+3. Perpetua item 1: "proposes 'ASIN mirroring'" → Perpetua names ASIN mirroring as a **trap to avoid** (the same product in the top organic slot and the top-of-search ad) and says branded ads should "expand coverage rather than duplicate it".
+4. §I.3 Amazon's FTC response: winning bids fell 50% "2019 to 2024/25" → "2019 to 2025". Conversion "up more than 24%" → "grew 24% from 2021 to 2025". The quote "generalized 'second-price' auction" → "generalized second price auction dynamics". Added Amazon's own mechanism sentence (a winner below the soft reserve "pay[s] their bid") and the "128th bid" figure. §I.4, stable item 12 and P-23 were updated to match.
+5. §D.6, §G.3, stable item 7, P-13: "organic and paid combined" was stated as an Amazon rule → it is Trellis's reading, tagged [opinion]; Amazon's schema says only that impressions "originate from the search results page". The SQP `asin` report option was called an option → it is **required** (space-separated, 200-character limit), so the SP-API report is ASIN-scoped.
+6. §D.1, stable item 5, P-9: the STIS denominator was "all ad impressions for that term" → the impressions of one ad product on that term across advertisers (Amazon's example: "20% of all Sponsored Products ad impressions"). The seat's three-denominator note was adjusted to match.
+7. §A.5, P-20: `KEYWORDS_RELATED_TO_PEER_BRANDS_PRODUCT_CATEGORY` was called a "Sponsored Brands theme" → it is in the v1 model's shared `ThemeMatchType`/`MatchType` enums, and no page read says which ad product accepts it (new Open question 15).
+8. §C.1: the 2024-06-13 note "introduced the goal" → the goal predates it (2023-11-06 note: "recently launched"; 2023-12-26 note: video under it).
+9. §C.4: budget-rule `IS` was "impression share" → "top-of-search impression share (`metricName` is `IS`)".
+10. §B.5, Open question 3, new P-39: the marketplace conflict now includes the RSOV what's-new page (modified 2026-10-09), which lists Egypt and not Ireland.
+11. §B.6: formats quoted verbatim; added Brand Gallery (an RSOV-only format, 2026-05-28) and the unread RSOV "additional terms".
+12. §B.8, Open question 2, run-time: the validation answer was "only validity and a reason" → it also returns a required `keywordReservationValidationId`. The behavioural sentence "Treat as a write until Amazon says otherwise" was rewritten as facts.
+13. §B.3: "cannot sit in more than one RSOV campaign" → verbatim "multiple campaigns or ad groups".
+14. §B.1: the `BRAND_IMPRESSION_SHARE` description is quoted verbatim ("grown [sic]"), with the IE error code `GOAL_BASED_CAMPAIGNS_DISABLED` added.
+15. §E.4: added that SB's up-to-300% bid adjustment applies only to "Drive page visits" campaigns.
+16. §E.1: "Product targeting takes any ASIN, own or other" is not on GB2JECV9CJK6R6AL → reworded to what that page says, plus the pages that do say it.
+17. §F.1: the change log has entries "for March and April 2026" → "October 2025 through April 2026".
+18. §F.3: added the healthcare "(excluding personal care products)" scope and the packaging exemption. §F.5: the Store-title rule quoted verbatim ("Sale," "Discount," "Deals").
+19. §F.7, P-33: removed "amended opinion" and "804 F.3d 930" (not on the page read). 1-800 Contacts: "14 rivals" → "at least 14"; "docket 18-3848" → that is the Second Circuit appeal number, and the FTC docket is 9372; the ALJ date is October 20, 2017.
+20. §J.1: Ad Badger's claim that prompts became billable on March 25, 2026, which the file said had no Amazon page behind it → confirmed by Amazon's prompts launch page (GA and CPC charging from 2026-03-25; prompts also appear on detail pages). The 2026-08-24 note covers `spPromptAdExtension` and `spVideoAdExtension`, not SB. The 2025-07-15 Alexa+ note is titled a test. New P-38 and stable item 18; Open question 9 and Ad Badger item 15 were updated.
+21. §D.2: v3 TOS IS report types made exact (`spCampaigns`, `sbCampaigns`, `sbTargeting`, plus the SP targeting report).
+22. §D.4: added that the SP keyword-recommendations guide asks v3/v4 integrators to move to v5.
+23. §D.7: the CBR is marked SB-only in the help table. The benchmark-brand definition is now cited to its dimension page, not the cross-program report page.
+24. §D.5: the partner-only claim is now sourced to the Partner Opportunities overview.
+25. §G.1: added the exact SQP/SCP availability sentence and the Search Terms report's ~12 million item limit.
+26. Ad Badger item 4: "almost 1 in 3" was the share of sellers "with this problem" → the share "starting to implement this strategy" (segmentation).
+27. Trellis item 7: "AI share of voice (Rufus and similar)" → "which brands ChatGPT, Claude and similar assistants recommend". Page date added (2026-10-05).
+28. m19 item 5: "undated" → datePublished 2026-10-27. Intentwise items: filters quoted verbatim and dates added. Haus: "Google text-ad campaigns exclusively targeting brand-name keywords" added. AmazOwl: context "Product Display Ads" added. Perpetua item 2: quotes made exact.
+29. Helium 10 row: "nothing on ad slots" → it places "Sponsored Products Related to this Item" above customer Q&A. Added as item 11a and to the layout disagreement.
+30. Blake, Nosko & Tadelis, P-26: "on Bing and Yahoo… clicks moved to eBay's organic link" is not in the abstracts read → cut back to the abstract (plus Coviello et al.'s one-line summary of it).
+31. Coviello, Gneezy & Goette: "a site whose brand and category queries face heavy competition" (unsourced) → the abstract's own framing, "paid links to branded search terms".
+32. Golden & Horton, P-28: "online labour market" (not in the abstract) → Google search ads suspended in random US markets.
+33. Simonov & Hill, P-27: "removed from the top position" → "top paid position"; "bounced back" → "returned to Bing".
+34. Dash et al., P-31: "lower-rated" → the abstract's "of poorer quality".
+35. Lewis & Nguyen: "in three advertisers' campaigns" added. Joo, Shi & Abhishek: the mechanism and simulation added. Farronato et al.: "browser-extension" dropped (not on the NBER page), revision date added. Moshary: the transfer was rewritten to what the abstract supports. Chiou & Tucker: the transfer now says the study is about resellers of the same product, not rival brands.
+36. Stable item 6, P-12, §G.2: "the only Amazon-sourced competitor identities" → SB brand recommendations by keyword also name brands (without shares).
+37. Added §E.8a, P-37, stable item 17 and Open question 16: the SB v4 API's own statement that "targeting your own brands usually results in lower performance than targeting competitors' brands", with no metric named. It was also added to *Where practitioners disagree*.
+38. Run-time: "The seller's own brand query is the row to read first" (a sequence) → a factual sentence. Added the SB negative-brand recommendations read and the prompt reports.
+39. Page names corrected in text and table: *What is Keyword Targeting? How It Works, Benefits, Examples*; *Your complete guide to Sponsored Brands*; *How to set up Amazon Sponsored Brands reserve share of voice*; *Introducing the search term impression rank report*; *Top-of-search impression share (top-of-search IS) to measure brand awareness*; *A complete guide to Sponsored Brands and display ads moderation*; the Brand Registry page title; the FTC November title ("Auction-Rigging"). Ten verifier-read pages were added to the table.
+40. §J.4 added: unBoxed Toronto 2026 on homepage-hero share-of-voice packages.
+
+**Removals.** "804 F.3d 930" and "amended opinion" (MTM); "on Bing and Yahoo" and "the clicks moved to eBay's organic link" (Blake et al.); "online labour market" (Golden & Horton); "heavy competition" (Coviello et al.); "lower-rated" (Dash et al.); "browser-extension" (Farronato et al.). No position was removed outright. Two behavioural sentences outside the positions were rewritten as facts (run-time "row to read first"; Open question 2 "Treat as a write"). One unsourced claim moved to the seat's list (ASIN-valued SP search-term rows as product-page matches). The positions were checked against the wording rule, and none states when to ask, a sequence, a threshold as a rule, a ceiling or how cx talks. P-34 (Perpetua's 40%) is stated as Perpetua's.
+
+**Stayed unreachable.** The verifier re-tried these: EUR-Lex *Google France* (HTTP 202, empty body); Search Engine Land *The data behind incrementality on Amazon* (403 to curl and WebFetch); the VoxEU column (403 to curl and WebFetch); the Justia MTM opinion (403); the Seller Central SQP help page (a JavaScript shell to GET; its content call is a POST that is out of bounds). Not re-tried, so still as the research seat recorded them: the Wilson Sonsini and IP Update notes; the govinfo 1-800 Contacts PDF; the FTC press-release search page; the Amazon Ads blog page that served a landing page. Not opened by either seat: the RSOV "additional terms".
+
+**Hosts.** The verifier called no Amazon Ads API host (advertising-api*.amazon.com), no Login with Amazon endpoint, no Amazon Ads MCP server endpoint, no SP-API host, and no Seller Central `mons-api` POST. It wrote nothing but this file and touched no git repository. Some docs pages show `advertising-api.amazon.com` inside sample curl commands; those samples were read as text and never executed.
+
+### Gap fill 2026-11-09
+
+Done by the research seat on 2026-11-09, after the verification above, for three completeness critics' two gaps and six contradictions. Every page cited below was opened by plain GET (curl, or the WebFetch tool where curl met a challenge); help pages were decoded from `window.__scV2Content` and their `lastModified` read; API models were parsed as JSON or YAML. No Ads API, Login with Amazon, Amazon Ads MCP or SP-API host was called; no Seller Central `mons-api` POST was made; nothing was written outside this file, and no git repository was touched. Sample curl commands inside docs pages were read as text only.
+
+**Gaps filled (2).**
+
+1. **Competitor stock-outs and Featured Offer loss as conquest moments.** Added §K (five items: the Featured Offer gate verbatim from *Understand Featured Offer eligibility*, lastModified 2026-10-16, including "impressions for the previous holder stop immediately" and the SKU rule; the out-of-stock pause from the library guide and the FAQs; the seat's mechanism for a rival's ads; the reverse case of a reseller holding one's own Featured Offer, in which SP and SD stop and SB keeps serving; and the fact that no Amazon Ads read reports a rival's stock). Added practitioner items 16–23 (Perpetua Competitor Signals, Jarvio, Netpeak, Eva, Trellis, Canopy, Pattern UK, a seller in the Seller Forums) and a disagreement bullet: Trellis's Qore page says SP ads "keep running" after a Buy Box loss, which Amazon's help page contradicts. Added research items 17a–17b (Conlon & Mortimer 2013; Musalem et al. 2010), with the note that no study of advertising during a rival's stock-out was found. Added the run-time subsection *A rival's stock and Featured Offer* with Keepa fields read from Keepa's public structs (`buyBoxSellerIdHistory` with "-1"/"-2", `stats.buyBoxStats.percentageWon`, `stats.buyBoxIsUnqualified`, price `-1` = no offer, `outOfStockPercentage30/90/InInterval`, `outOfStockCountAmazon30/90`, `availabilityAmazon` codes, `offers[].stockCSV`) and SP-API Product Pricing reads parsed from the models (`getItemOffers` with `IsBuyBoxWinner`, `getCompetitiveSummary` with `featuredBuyingOptions` and `similarItems`). Stable items 19, 20, 26; positions P-40 to P-46, P-58; seat positions; Open questions 18 and 22.
+2. **Sponsored Brands video and Store Spotlight as conquest and defence formats.** Added §L (seven items: the creative types and formats from the SB 4.0 `CreativeType` enum and the *Ad types overview*, *Video ad* and *Store spotlight* guides; placements by goal and format from *Campaign goals*, *Understand Sponsored Brands*, *Get started with Sponsored Brands video* and the ad-specs page; Amazon's own suggestion of "brand names that are similar to yours" as video keywords and video's "own ad placement and auction"; the moderation guide's rule that videos follow chapters 1–5, chapter 4 forbidding competitor names; BTR Media's *Ad Intel* column on video for competitor shoppers; Amazon's 2023 figure that 60% of brand searchers bought a competitor within 30 days; and the SB video report columns). Added practitioner items 24–30 (Skale Strategy, ATTN Agency, AMALYZE, Amplisell, AMALYTIX, Perpetua 2021, and the seat's note that no method-stated measurement of SB video on competitor terms or ASINs was found). §H.1 now carries the help page's "at least 4 pages" beside the API's "at least three subpages". Stable items 21–23; positions P-47 to P-53; seat positions; Open questions 19–21.
+
+**Contradictions settled (6).**
+
+1. **Reserve share of voice figures (owner: this file).** Re-opened the three help pages, the library guide and the what's-new page. The conflicts stand as Amazon's own: 15–92 days and $3,000 (G86SD7HK6NHHRB9B, lastModified 2026-07-07), 30–92 days (GXXPFXLSVX9VD3U9, lastModified 2026-08-11), "e.g., 14–90 days" and $6K US (library guide, modified 2026-10-09), Ireland in the library guide's list and Egypt instead in the what's-new page's (also modified 2026-10-09). Corrections here: "not available in Spain" is on **both** help pages G86SD7HK6NHHRB9B and GXXPFXLSVX9VD3U9 (this file had it on one; §B.4, §B.5, P-39, Open question 3 fixed); the keyword count 5–1,000 is the one figure all three help pages share (§B.5, P-6). `areas/bids-and-placements.md` P-33, as it reads on 2026-11-09, already carries the flight, minimum and marketplace discrepancy and points here, so it agrees with these sources; `areas/keyword-and-product-targeting.md` P-41's keyword count alone is consistent with all three help pages.
+2. **`/v2/stores` (owner: `method/ads-api-surface.md`).** Re-read the *Deprecations* page: "V2 Stores | `/v2/stores` | /brand/stores/v1/storePages/list | January 2026 | June 1, 2026". Re-read the Brand Home model (`stores/list` returns `StoreInfo.brandEntityId`; `storePages/list` lists pages) and the SB 3.0 YAML (`GET /brands` returns `brandEntityId` for sellers). The Brand Store overview guide still shows `GET /v2/stores` (as `beyond/ad-products-and-eligibility.md` noted). ads-api-surface is right. §H.2 and the run-time "The brand and its ids" no longer send cx to `/v2/stores`; stable item 24 and P-54 added.
+3. **Which SB spec carries which negative path (owner: `method/ads-api-surface.md`).** Parsed the three models. The YAML "Amazon Ads API - Sponsored Brands" version 3.0 carries `/sb/negativeKeywords` and `/sb/negativeTargets` (and `/sb/recommendations/targets/brand`) and not `/sb/negativeTargets/brands/recommendations`; the JSON file titled "Sponsored Brands 3.0" (`SponsoredBrands_prod_3p.json`, 50 paths) and the SB 4.0 model (48 paths) carry `/sb/negativeTargets/brands/recommendations` (`SBTargetingGetNegativeBrands`) and not `/sb/negativeKeywords` or `/sb/negativeTargets`. So ads-api-surface §I is right about `/sb/negativeKeywords` on the YAML, search-terms is right about the JSON file, and keyword-targeting and this file are right that the brand-recommendation path is on SB v4; the files disagreed because two different files are both called "Sponsored Brands 3.0". §E.8, §E.8a and the run-time list now name the model for each path; P-55 added. (`areas/search-terms-and-negatives.md` P-55 states the same.)
+4. **`topOfSearchImpressionShare` report types (owner: `method/reporting.md`).** Re-read the v3 columns page ("**Report types**: spCampaigns, sbCampaigns, sbTargeting"), the *Targeting reports* page (SP and SB base metrics) and the *Campaign reports* page (SP "Group by `campaign`" and "Group by `campaignPlacement`"; SB "Group by `campaign`"). This file's statement stands and is made exact (§D.2, run-time, P-56); `method/reporting.md` §C's list of `spCampaigns` and `spTargeting` only is incomplete against these three pages.
+5. **Conversion rate by ad position (owner: `areas/bids-and-placements.md`).** Re-read the four Crossref abstracts. Each file's single citation is accurate for its paper: Ghose & Yang, "conversion rates are highest at the top and decrease with rank" and "profits are often higher at the middle positions" (this file's quote is verbatim); Rutz, Bucklin & Sonnier, "higher positions increase both the click-through and conversion rates"; Agarwal, Hosanagar & Smith, "conversion rate increases with position" (lower down); Narayanan & Kalyanam, naive comparisons "biased due to these selection issues", causal effects "much smaller" and only at some positions. `areas/bids-and-placements.md` P-51 reconciles them. Here: research item 14 now says this file uses Ghose & Yang for profit only and points to bids P-51; new item 14a adds Narayanan & Kalyanam, including the moderator that position effects are "weaker when the keyword phrase has specific brand or product information", which bears on brand queries; stable item 25 and P-57 added.
+6. **Golden & Horton and `method/change-measurement.md` (owner: change-measurement).** Re-read the Golden & Horton abstract ("only 63% as many as a nonexperimental estimate would have suggested") and change-measurement, which cites neither Golden & Horton nor any finding that a naive pause read overstates the loss; its one directional pause claim is Ju, Zhao & Aral's, that a short pause understates a loss that builds through rankings. This file's sentence "`method/change-measurement.md` covers why a before/after read of a pause overstates the loss" was wrong and is replaced (research item 5): the overstatement is Golden & Horton's own finding for an unspecified nonexperimental estimate, change-measurement carries the opposite-signed finding from a different setting, and neither settles the sign for Amazon.
+
+**Declined (0).** Both gaps were right on the facts: Amazon's Featured Offer page does state that SP and SD stop and SB continues when another seller holds the Featured Offer, and Amazon's moderation guide does bind videos to the no-competitor-name rule. Neither gap was declined. One part of gap 2 could not be filled as asked: "practitioner measurements of SBV conquest CTR and CVR" with a stated method were not found; what exists (single-client and portfolio claims, and Amazon-published advertiser-provided data) is recorded as such in items 24–30 and P-52.
+
+**Unreachable in the gap fill.** Pacvue blog posts (Cloudflare, then 403); LandingCube and Seller Metrics (bot verification); Search Engine Land (Cloudflare); Seller Central Featured Offer help pages (JavaScript shell; the content call is the forbidden POST). Listed under *Unreachable*.
+
+**Pending.** The gap-fill additions above have not been re-checked by an adversarial verifier: Pending verification.