We use optional cookies to analyze traffic and personalize content.
Ads turn over in days.
Storefronts stand still.
In September, 58% of the stores we track ran ads, and the typical Meta ad that ended had run for just 6 days. The storefronts those ads point to moved far less: of 1,287 we scored for AI-agent readiness, not one earned an A. Express wallets sit on about 85% of stores; buy-now-pay-later on fewer than 3 in 100.
01 Advertising who ran ads, and for how long
Most brands advertise. Few let an ad live long.
Ads on Meta and Google whose delivery window overlapped September, counted only when the advertiser is verified as the brand. 179 tracked stores checked in September
18% of Meta ads carried an offer, and three in four advertisers ran at least one. The median advertised discount was 25%, with discounts and free shipping the most common hooks. In Europe, where Meta discloses targeting, Germany and the UK were the most-targeted markets, and over half of ads aimed at a single country.
02 AI readiness can a shopping agent use this store
Not one storefront earned an A.
Stores scored 0-100 on how machine-readable and agent-ready they are: crawlability, structured data, llms.txt, entity signals. 1,287 stores scored in the 60 days to 30 September
Checkout rails are increasingly handed out by the platform. What stays in the brand's hands is being found, read and trusted by an AI assistant, and on that score the market is still in the D band.
03 Checkout the wallets a machine can tap
Express checkout is the floor. Buy-now-pay-later is still rare.
Wallet, express-checkout and buy-now-pay-later methods detected on storefronts on 30 September. Card networks are excluded. % of 684 stores with a detected payment method
The most common buy-now-pay-later method, Zip, appears on 2.5% of stores, and Klarna on 1.9%. The top five wallets are the same, in the same order, as at the end of August.
04 Fulfillment who actually ships the order
Everyone takes the payment. Far fewer say who delivers.
Consumer-facing shipping carriers detected on 30 September, on stores that expose one at all. % of 195 stores with a detected carrier
Only 195 stores exposed a carrier we could read. Delivery is where storefronts go quiet, so an AI assistant asked "who brings it, and when" usually cannot answer. The stores that publish it win that question by default.
05 New stores where the next brands are landing
New stores launch small, and mostly on .com.
Domains chosen by new Shopify storefronts discovered in September, and how existing catalogs moved. 812 new stores sampled
The median new store opened with just 9 products, and 44% opened with 3 to 10. Established catalogs were balanced: of 64 stores whose product count changed, 33 grew and 30 trimmed, with a median change of +1 product.
06 Trust is the listing even real
Most catalogs are clean. The noise sits in a few.
Suspected decoy listings - shown but never really purchasable, price-flickering, or orphaned - from catalog scans run in September. One store with more than half of its catalog flagged is excluded as a likely detection error. 289 active stores with more than 10 listings
We call these suspected listings, not deliberate honeypots: most are catalog debris. For an AI shopper, though, a dead listing is a dead recommendation.
· How this is built why you can trust the numbers
Not a survey. A month of continuous, public-signal monitoring.
2,088 stores, September only
Changes logged between 1 and 30 September 2026. No later data is mixed in.
Every figure shows its n
A store counts only for subjects it had measurable data for. No data means excluded, never a zero.
Public information only
Storefronts and public ad libraries. No private, gated, personal or account-restricted data.
100% anonymized
Aggregated across the market. No individual brand is named, shown or identifiable.
→ Your move
You just saw six cuts of the September market. Where does your store sit?
This is a teaser. The full September Radar covers AI readiness, advertising, assortment, what the market sells, new stores, programs, global reach, platforms, payments, fulfillment and decoy listings, with the sample size and method behind every figure.
Based on RivalSweeper's daily monitoring of 2,088 ecommerce stores throughout September 2026. Figures are aggregated across the market; no individual brand is named or identifiable. Every metric reports its own sample size. Payment, shipping, advertising and suspected-decoy figures are estimates from public signals and may be revised. All figures are provided "as is" for informational purposes only, and are not financial, legal, or investment advice. This report includes AI-assisted analysis. Spotted something off? Tell us at [email protected]. Copyright © 2026 RivalSweeper. All rights reserved.