Ecommerce Weekly | Prime's fall event rehearses the peak as product and price signals flow to AI channels
October 5–11, 2026: Prime Big Deal Days data, Meta's AI ad tools, Google's UCP integration hub, compare-at prices for agents, and Merchant Center image rules
The week combined a real sales event with infrastructure for AI shopping channels. Prime Big Deal Days produced the closest market-wide reading before the peak. At Advertising Week, Meta opened new-versus-existing-customer controls and static-to-video generation to all advertisers while testing an assistant that changes budgets and targeting. Google rolled out the Merchant Center UCP integration hub in the US. Shopify began sharing compare-at prices with AI shopping agents. Merchant Center stopped blocking products for low image quality, while a 500 × 500 pixel minimum is due at the end of January 2027. Official statements and third-party data are kept separate, and actions are limited to what can be finished and read back before BFCM.
Editorial linePublicThe week's main story: a sales rehearsal plus a set of AI-channel interfaces
Open the editorial line and its boundaries
Prime Big Deal Days on October 6–7 was the last large promotion before the peak. Adobe data put US online retail spending over the two days at $9.86 billion, up 8.5% year over year, with mobile at 52%. That is market-wide, not Amazon-only, so for an independent store it works as a rehearsal: traffic prices, discount depth, and mobile checkout return at larger scale in November.
In the same week, three platforms pushed decisions and transactions toward AI channels. Meta opened new-versus-existing-customer controls and static-to-video generation to all advertisers and began testing an assistant that changes budgets and targeting. Google gradually rolled out the Merchant Center UCP integration hub in the US, where merchants can offer cart transfer without handing over checkout. Shopify shared compare-at prices with AI shopping agents, so agents can see how deep a markdown is.
What these interfaces share is that platforms now read products, prices, and carts directly. An independent store should first confirm the data it hands over is correct, then decide whether to connect more.
Verified facts
- October 6–7: Adobe counted $9.86 billion in US online retail spend, up 8.5%; mobile 52.0%; BNPL 6.0% of orders; discounts peaked around 20% off list.
- October 6: Meta opened Customer Lifecycle Strategy to all advertisers, made static-to-video in Advantage+ creative generally available, and expanded the ads MCP server; assistant actions such as budget and targeting changes remain in testing.
- Around October 6: Google began rolling out the Merchant Center UCP integration hub in the US, supporting cart transfer and native checkout on Google, with identity linking marked "coming soon".
- October 5–9: Shopify Catalog started returning compare-at prices to AI shopping agents as list_price; Shop can recognize buyers while browsing and sync carts across devices.
- Around October 11: Merchant Center help says products with low-quality images continue to show; all product images must be at least 500 × 500 pixels, enforced from January 31, 2027.
Several of this week's capabilities are not settled yet: accounts of who can join the UCP hub differ, Meta's assistant actions are still in testing, and Google Ads automated promotions switch on by default from today, October 12. Fix the data you hand to platforms before deciding what to connect.
Impact read
Macro view: none of this week's AI interfaces brings traffic directly, but each shapes how AI channels understand your products. Compare-at prices, carts, customer segments, and product images used to serve page display; now they are also inputs to agents and ad systems. The most useful pre-peak work is data hygiene: truthful price indications, reliable customer lists, and compliant product images.
1. Prime Big Deal Days: $9.86 billion in US online spend over two days, the last rehearsal before the peak
Amazon announced on September 14 that Prime Big Deal Days would run October 6–7 in 22 countries, with dates varying in some. On October 8, Digital Commerce 360 reported Adobe data: US online retail spending over the two days was $9.86 billion, up 8.5% year over year (from $9.09 billion in 2025). Mobile accounted for 52.0%, about $5.1 billion, up 12.8%. Buy now, pay later enabled 6.0% of orders and $692.3 million in sales. Discounts peaked at about 20% off list, deepest in computers, TVs, electronics, toys, and apparel.
These figures are Adobe's view of US online retail as a whole, including retailers other than Amazon; Amazon did not publish its own sales total in that report.
What this means
Impact read
This is a market thermometer, not Amazon's scorecard. Growth of 8.5%, mobile above half, and discounts around 20% suggest shoppers were willing to buy early in October without discount depth running away. For an independent store, the useful comparison is not the total but your own traffic, ad cost, and conversion in the same window.
What it means for an independent brand
Independent stores that did not join Amazon's event were still affected: competitor markdowns, more price-comparison traffic, and higher ad auctions. Using your October 6–7 data as a rehearsal can expose mobile checkout, inventory, and support bottlenecks before November.
Decision: medium-to-high data confidence (one data provider, reported by trade media); its value for independent stores is rehearsal, not forecast.
Not yet
Avoid this week
- Do not set Adobe's market-wide growth as your BFCM target.
- Do not bolt on a new payment method before the peak because BNPL reached 6%; check your own order value and refund flow first.
Decision boundary
Applies to independent stores that ran October promotions or face Amazon price comparison. Impact: medium. Confidence: medium to high on the data. The judgment changes if Adobe revises the figures, or if your own data for those two days looks no different from normal (your category is not sensitive to Amazon events). It shows early-October buying intent; it does not prove November will grow at the same rate.
Micro action: export October 6–7 in store time alongside the same weekdays in the previous two weeks, and compare traffic, mobile share, CPC or CPM, conversion, and order value. If mobile conversion trails desktop clearly, schedule a mobile checkout review for next week, starting October 19.
2. Meta at Advertising Week: new-versus-existing customer controls for all, assistant budget changes still in testing
In its October 6 Advertising Week New York post, Meta said Customer Lifecycle Strategy is available to all advertisers, with acquisition campaigns receiving recommendations to exclude existing customers. Advantage+ creative can generate video from existing static assets and is generally available. The ads MCP server added creative upload, ad previews, performance tracking, shared drafts, and permission-gated actions, with ten partners integrated. Ads Creative Studio's creative insights are rolling out to more advertisers, and catalog product-video generation is in beta for eligible advertisers.
Still in testing: the Meta AI business assistant creating campaigns, generating creative, and changing targeting and budgets by conversation; scheduled tasks and alerts; direct checkout with Shopify in Meta Business Agent on Messenger. AI audience discovery in Detailed Targeting is due by the end of the year. Meta says advertisers who adopt the assistant's recommendations typically see 5% more conversions at the same cost per conversion; that is Meta's own figure.
What this means
Impact read
The most directly useful piece for independent stores is the new-versus-existing split: acquisition budget spends less on people who already bought. The rest points to "AI drafts, humans approve": MCP lets outside tools save changes as drafts, and the assistant's execution actions are still tests.
What it means for an independent brand
Customer Lifecycle Strategy needs reliable existing-customer data; stale or partial lists distort the exclusion. Static-to-video lowers creative cost, but each generated asset still needs checking against product facts.
Decision: the customer split is a high-certainty improvement to use now; treat assistant and MCP execution as read-only or draft-only.
Not yet
Avoid this week
- Do not let the assistant or MCP tools change budgets and targeting directly during the peak; keep them at drafts with human approval.
- Do not treat Meta's "5% more conversions" as your own account's expectation.
Decision boundary
Applies to independent stores advertising on Meta. Impact: medium. Confidence: high on the features, medium on commercial effect. The judgment changes if new-customer cost rises after excluding existing customers while new-customer volume does not grow, or if generated video often gets product facts wrong. It proves the tools are available; it does not prove they add incremental results in your account.
Action this week: prepare a refreshed existing-customer list for Customer Lifecycle Strategy, then list which automatic enhancements are on for live Advantage+ ads (background changes, text translation, AI-generated images, and similar). For creative that states product facts, sizes, or claims, switch off options that can rewrite content or assign a reviewer to each asset. There is no time for rework once the peak starts, so confirm generated content is accurate before November.
3. Google's Merchant Center UCP hub: offer cart transfer without handing over checkout
Search Engine Land reported on October 6, followed by ALM Corp on October 9, that Google began gradually rolling out the Merchant Center Universal Commerce Protocol (UCP) integration hub in the US. Google's developer guide lists three capabilities: transfer to merchant cart, where shoppers add items on Google and check out in the merchant's own cart, requiring the Cart API with a shorter validation path; native checkout on Google, where shoppers buy without leaving Google, requiring three core REST endpoints, order-status webhooks, and agreement to the UCP Program Addendum; and identity linking, marked "coming soon". Merchants publish a UCP profile at /.well-known/ucp whose declared capabilities must match their selection, and can test in a sandbox first.
Google's guide says Australia and Canada are expected early next year. Google's announcement page says "generally available for US selling merchants", while the developer guide and trade coverage say "gradually rolling out", and ALM mentions an interest form and selection, so accounts differ.
What this means
Impact read
Joining Google's AI shopping checkout used to mean handing checkout to Google. There is now a lighter path: hand over only the cart, while checkout, payment, and the customer relationship stay in your own store.
What it means for an independent brand
For independent stores selling in the US, cart transfer is a low-cost pilot. Merchants on Shopify and similar platforms will usually depend on whether the platform implements it for them rather than building endpoints. Sellers outside the US are not in scope yet.
Decision: the direction is clear while coverage and eligibility are still moving; US stores can assess cart transfer now and wait for more data on native checkout.
Not yet
Avoid this week
- Do not prioritize native checkout until you know how orders, refunds, and customer data are handed over.
- Do not change checkout or inventory systems mid-peak to join UCP.
Decision boundary
What is still missing is the data layer: Google has not said which source and performance data merchants will see for orders arriving through cart transfer, nor published attribution rules. Without them, a pilot can show only whether orders arrive, not what the AI shopping surface did for the decision.
Applies to US-market independent stores with development resources or platform support. Impact: medium. Confidence: high on direction, medium on eligibility details. The judgment changes if Google publishes data-sharing and attribution rules for transferred orders, or if platforms switch it on by default. Micro action: check whether UCP integration appears under Settings in Merchant Center, and ask your platform whether it implements cart transfer for you.
4. Compare-at prices reach AI agents: the size of the markdown is now machine-readable
Shopify's merchant and developer changelogs on October 9 said Shopify Catalog now shares the merchant's compare-at price with AI shopping agents, returned in the Catalog API as list_price and list_price_range, effective October 5. The field appears only when three conditions hold: the merchant shares compare-at prices through Catalog Mapping in the admin (on by default for most stores, off for stores that use compare-at as MSRP or another reference price); the compare-at price is higher than the price in that buyer's context; and the merchant shows compare-at prices in the buyer's region. Stores that hide them from EEA buyers do not return the field to those buyers.
In the same week, Shopify said Shop can recognize buyers while they browse, syncing carts across devices, enabling one-tap sign-in, and pre-filling Shop Pay (October 5). A developer changelog reminded that the deprecated Global Catalog REST API stops serving traffic on November 2 and must be replaced by the UCP-based Global Catalog MCP (October 9).
What this means
Impact read
Compare-at prices used to be a visual cue on the page; now they are an input to agent reasoning. Agents compare "was" and "now" and may tell the buyer how big the discount is. Whether the compare-at price is honest moves from a copy question to a data-quality question.
What it means for an independent brand
Stores with persistently inflated compare-at prices, or that use MSRP as compare-at, carry the most risk: agents amplify these "discounts", and markets such as the EU have specific rules on price-reduction claims. Shopify already separates stores that hide compare-at from EEA buyers, but other markets follow the admin setting.
Decision: certain; medium impact for stores relying on compare-at for promotions. Confirm the Catalog Mapping setting and compare-at accuracy before the sale.
Not yet
Avoid this week
- Do not raise compare-at prices right before the sale to manufacture discount depth.
- If compare-at represents MSRP or a reference price, do not let it enter agent channels as a "discount".
Decision boundary
Applies to merchants using compare-at prices for promotions whose products are in Shopify Catalog. Impact: medium. Confidence: high. The judgment changes if agent channels do not display discount information, or if Shopify switches sharing off by default. It proves agents can read the markdown; it does not prove agents will recommend your products more.
Micro action: check the toggle in the Compare-at section of Catalog Mapping; sample 20 products with compare-at prices and confirm each reflects a real recent selling price; for products where it means MSRP, switch sharing off or use another field. Finish by October 23 and freeze changes during the sale.
Sources
5. Merchant Center image rules: low quality no longer stops products from showing, a 500 × 500 minimum arrives at the end of January
PPC News Feed reported on October 11 that Google Merchant Center no longer stops products with low-quality images (blurry, pixelated, low resolution) from showing in ads and free listings. The Google help page we checked says "your product will continue to show in ads and free listings" and lists a new requirement: all product images must be at least 500 × 500 pixels, enforced from January 31, 2027. Since July 2026, images above the old minimum but below 500 × 500 receive warnings.
Trade coverage also says promotional overlays (prices, calls to action, shipping details) remain a strict disapproval reason. The help page we reviewed does not address promotional overlays directly, so Google's image policy text is the reference.
What this means
Impact read
Products will not suddenly stop showing for blurry images during the peak, which lowers display risk in the sale period. The rule is not looser, though; it is replaced by a clear size threshold that is enforced after the peak.
What it means for an independent brand
Stores with large back catalogs and mixed image sources can schedule image fixes for December and January rather than squeezing them in before BFCM. New sale products should meet 500 × 500 or more from the start.
Decision: certain (official help page plus trade report); near-term risk falls, and one bulk check is needed before the end of January.
Not yet
Avoid this week
- Do not replace main product images in bulk during the sale; it affects both review and click performance.
- Do not add price, discount, or shipping text to product images, even though the low-quality rule changed.
Decision boundary
Applies to all independent stores using Google Merchant Center. Impact: low to medium. Confidence: high. The help page shows no update date, so the timing of the change relies on the trade report. The judgment changes if Google enforces the 500 × 500 threshold earlier or postpones it.
Micro action: filter image-size warnings in Merchant Center diagnostics, export the list, sort by sales, fix the top 100 products by December 15, and finish the rest by January 20.
Watchlist: TikTok Shop US enforcement policy, Google automated promotions, and the end of a search update
TikTok Shop US Seller Enforcement Policy. The third-party site Get Live recorded on October 7 that the US Seller University listed a new Seller Enforcement Policy published October 6. It described Account Health Rating (AHR) point deductions with escalation at 150, 100, 50, and 0 points, fund holds of 45, 90, or 365 days, "connected account" treatment based on shared contact, payment, device, or business details, and a first appeal within 30 days of notice. We could not open the policy page to verify the original text, so this stays a single-source signal.
Google Ads automated promotions. Around October 5, Google told advertisers that from October 12 automated promotions would be on by default for Search and Performance Max campaigns with location assets and no promotion assets. That is this Monday; we will follow up on how it shows up in accounts next week.
Google September spam update. Converting the Search Status Dashboard duration, the update ended around October 7–8.
Impact read
What to do: brands selling on TikTok Shop US should open the policy text in Seller Center and check the connected-account and fund-hold clauses, especially for shops run by agencies. The other two items are handled in later issues.
The five changes on one operating chain: AI channels start reading your data directly
| Layer | This week's change | What it can answer | What it cannot prove |
|---|---|---|---|
| Demand and competition | Prime Big Deal Days market data | Early-October online demand and mobile share | That November grows at the same rate |
| Price signals | Compare-at prices in Shopify Catalog | Whether agents can read markdown depth | That agents recommend your products more |
| Cart and checkout | Merchant Center UCP integration hub | Whether shoppers can add on Google and check out in your store | How data and attribution for transferred orders are shared |
| Ad decisions | Meta customer split, assistant tests | Whether acquisition spends less on existing customers | That assistant recommendations are better |
| Product assets | Merchant Center image rules | Whether images can stop products showing in the sale | That new images raise click-through |
Impact read
Three of the five changes are platforms reading your data: compare-at prices, carts, product images. Two are platforms making decisions for you: the assistant and automated delivery. The first group risks amplifying inaccurate data; the second risks decisions without human approval. Before the peak, fix the data first and discuss automation second.
Counter-signals: the compare-at check finds many products whose "was" price cannot be supported; acquisition cost rises after excluding existing customers (a list-quality problem); no transferred orders appear after joining UCP; your Prime-window data looks no different from normal. If any of these appear, keep the conclusion local and do not extend it store-wide.
Baseline execution checklist: four optional actions, finish one first
Action card
Use the two Prime days as a sale rehearsal
Export October 6–7 in store time with the same weekdays in the previous two weeks; compare traffic, mobile share, ad unit cost, conversion, and order value; list three bottlenecks most likely to grow at BFCM.
- Applies to
- Independent stores with October promotions or Amazon price exposure
- Done when
- A comparison table exists and each of the three bottlenecks has an owner.
- Owner
- Growth lead
- Effort
- 60–90 minutes
- Stop when
- If another major campaign or a tracking outage overlaps the window, note it and draw no conclusion.
- Review at
- 10/16/2026, 2:00:00 PM
Related resources
Action card
Review Meta's AI creative enhancements
List live Advantage+ ads and record which automatic enhancements are on; for creative stating product facts, sizes, or claims, switch off options that can rewrite content or assign a reviewer.
- Applies to
- Independent stores advertising on Meta
- Done when
- Each live ad has a recorded enhancement setting and reviewer.
- Owner
- Paid media lead
- Effort
- 60 minutes
- Stop when
- If performance drops clearly after switching enhancements off, restore them and move to per-asset human review.
- Review at
- 10/19/2026, 2:00:00 PM
Related resources
Action card
Check compare-at prices and the Catalog Mapping setting
Confirm in the admin whether compare-at prices are shared with AI agents; sample 20 products with compare-at prices and confirm each is a real recent selling price; handle MSRP-style compare-at separately.
- Applies to
- Shopify merchants using compare-at prices for promotions
- Done when
- All 20 products have a check record and the toggle state is screenshotted.
- Owner
- Pricing or merchandising lead
- Effort
- 60 minutes
- Stop when
- For products whose "was" price cannot be supported, remove the compare-at price and do not use it in the sale.
- Review at
- 10/23/2026, 2:00:00 PM
Related resources
Action card
Export Merchant Center image-size warnings
Filter warnings for images below 500 × 500, sort by sales, and schedule fixes for December and January.
- Applies to
- Stores using Google Merchant Center
- Done when
- A sales-ranked fix list with completion dates exists.
- Owner
- Product operations
- Effort
- 30 minutes
- Stop when
- Do not replace main images in bulk during the sale; only log them.
- Review at
- 10/16/2026, 2:00:00 PM
Related resources
Priorities at the end: fix the data you hand to platforms first
| Priority | Who should look first | Action | Why |
|---|---|---|---|
| P1 | Review | Teams advertising on Meta | Review Advantage+ creative enhancements by October 19 | AI-generated or rewritten creative must match product facts, and there is no time for rework during the peak |
| P1 | Check | Shopify merchants using compare-at prices | Check compare-at accuracy and the Catalog Mapping toggle | Discount depth now flows directly to AI agents |
| P2 | Rehearse | Every independent store preparing BFCM | Use the two Prime days to find three bottlenecks | It is the closest real reading before the sale |
Avoid this week
- Do not let the assistant or MCP change budgets and targeting directly during the peak.
- Do not raise compare-at prices before the sale.
- Do not change checkout systems mid-peak to join UCP.
- Do not replace main product images in bulk during the sale.
Sign in to continue with the analysis and execution cards available to your plan
You are viewing the Public edition. The server returns the sections available to your account, while unopened content stays out of public APIs and page source.
View plan access