Photo: Pexels
For years, the advice given to online sellers was simple: rank on Google, run good ads, and keep your storefront tidy. That advice is not wrong. It is just incomplete now, because a new kind of shopper has walked into the store — one that never browses the way a human does.
On 29 October 2026, Adobe published data that should change how every small seller thinks about the coming holiday season. Shoppers who arrived at United States retail websites after researching a product through an AI tool generated 83% more revenue per visit than shoppers from other channels in September 2026, and they converted 90% better. A year earlier, the same channel was worth roughly the same as any other visit — and in August 2025 it actually converted 9% worse. In about fourteen months, AI-referred traffic went from the weakest visitor on the site to the most valuable one.
That is not a curiosity. It is a shift in who decides what gets bought, and it arrives weeks before Black Friday, the heaviest shopping stretch of the year. Sellers who understand it now have a short window to prepare. Sellers who wait will find that their products were never even shown to the customer.
The numbers behind the shift
Adobe's figures come from a very large sample: more than one trillion visits to US retail sites, 100 million products, and 18 product categories, plus a survey of more than 5,000 US consumers. The September 2026 numbers are worth stating plainly, because each one points to a different action.
First, volume. AI traffic to US retail sites grew 71% year over year in September, and it has grown every month since Adobe began tracking it in October 2024 — a cumulative rise of 1,772%. This is no longer an experiment running at the edge of the market.
Second, behaviour. Once they land on a retail site, AI-referred visitors spend 73% more time on site, view 25% more pages, and add products to their carts at a 53% higher rate than other visitors. They are not bouncing. They are checking.
Third, value. The 83% revenue-per-visit advantage in September was the largest gap Adobe had recorded in 2026, and Adobe expects that advantage to double during the holiday season on a year-over-year basis. The reason is straightforward: these shoppers tend to arrive after the comparison work is already done. They are not starting their research on your product page. They are finishing it there.
Fourth, the holiday outlook. Adobe's September forecast put total US online holiday sales (1 November to 31 December) at $275.1 billion, up 6.7% on 2025, with Cyber Monday at $15.1 billion and Black Friday at $12.9 billion. AI traffic is expected to rise 130% across the season, with spikes of 159% on Thanksgiving, 95% on Black Friday, and 88% on Cyber Monday. In Adobe's consumer survey, 77% of people who had used AI for shopping said they felt more confident in their purchase, and 69% said they were less likely to return what they bought. A confident buyer who rarely returns things is exactly the customer a small seller wants.
Why AI shoppers buy differently
A traditional shopper types a few words into a search engine, scans ten blue links, opens three tabs, and compares prices by hand. An AI shopper asks a question in plain language — "best wireless earbuds under $50 with long battery life" — and receives an answer, not a list. The AI tool has already filtered, ranked, and summarised. By the time the shopper clicks through to a store, the shortlist is set.
That changes the job of a product listing. In the search era, a listing had to attract a click. In the AI era, a listing has to survive being read, understood, and quoted by a machine before any human ever sees it. If the AI tool cannot extract a clear price, a clear set of specifications, a clear shipping and returns policy, and a clear reason why this product fits the question, it moves on to a rival listing that does offer those facts. The shopper never knows your product existed.
This is also why AI shoppers convert so well. They arrive pre-sold on a category and often on a specific product. The remaining questions are practical: is it in stock, what does delivery cost, how fast does it arrive, and what happens if it breaks. Stores that answer those questions on the page close the sale. Stores that bury them behind vague marketing copy lose a buyer who was ready to pay.
The visibility problem: more than a third of homepages are unreadable
Here is the uncomfortable part of Adobe's data. The company built a Citation Readability Score, which measures how much of a web page AI systems can actually understand and cite. In September 2026, the average US retail homepage scored just 63%. More than a third of homepage content was invisible to the systems sending retailers their best customers.
The gap between categories is wide. Electronics led with an overall score of 79%, helped by category pages full of setup guides, troubleshooting articles, and detailed specifications. Grocery followed at 55%. General merchandise homepages scored only 40%, because navigation-heavy, promotion-led layouts give an AI tool very little to work with. The pages that scored best were the ones that explain: product details, buying guides, frequently asked questions, and return policies.
Put bluntly, most stores are currently built for human eyes and search-engine crawlers, not for the answer engines that now sit between the shopper and the store. A beautiful homepage banner means nothing to a system that reads text. A product page with three photographs and a two-line description gives an AI tool almost nothing to quote — so it quotes someone else.
Shopify's shop.md: the personal shopper arrives
The timing matters because the tools that turn AI assistants into actual buyers landed in October. Shopify's shop.md, reported on 19 October and folded into the company's Fall '26 Edition on 28 October, turns an AI agent into a personal shopper. A shopper points an agent such as ChatGPT or Claude at shop.md, connects a Shop account, and the agent can search more than one billion Shopify products, use the shopper's sizes, preferences, and order history to narrow the choice, check out with Shop Pay, and track the order — all inside the conversation.
Two details deserve attention. First, shop.md is a shopper-side tool, not a store setting. Sellers do not install it; shoppers bring it with them. Second, what determines whether a product appears is catalogue eligibility: complete product data, clear policies, and correct shipping configuration. Shopify's Fall '26 Edition added a catalogue eligibility view in the admin and a knowledge-base page where merchants publish the facts agents receive — return windows, shipping destinations, gift services. It also extended Shopify connectors to nine agents, including ChatGPT, Claude, Gemini, Copilot, Grok, and Perplexity, and added selling inside Google Search's AI Mode and the Gemini app.
The message for sellers on any platform is the same, whether they use Shopify, a marketplace, or their own website. The agent shops from structured facts. Thin product data gets skipped. That warning is already circulating among sellers on social media: an Instagram reel from @shoperatorai on 28 October put it in exactly those terms, noting that agents "never visit your site" in the old sense — they query who has the item in stock at the right price — and that incomplete sizes, materials, and stock information means a listing is passed over. A Threads post by Muhammad Umair Amjad on 29 October made the complementary point, citing Adobe's 83% figure to argue that marketers should judge channels by commercial value, not by raw session volume.
What this means for marketplace sellers
Sellers on marketplaces such as eBay and Amazon face a version of the same shift, with one difference: they do not control the platform's pages, so their leverage is the listing itself. Titles, item specifics, descriptions, and policies are the data the AI reads. A title stuffed with keywords but missing the one attribute the shopper asked about — size, material, compatibility, wattage — is a title an AI answer will skip. A description that states the facts in plain sentences will be quoted. A description written only to sound persuasive will not.
There is a second, quieter implication. AI tools compare relentlessly. A shopper who asks for "the cheapest reliable option" will be shown the cheapest reliable option, and a seller competing only on price will be pushed into a race they cannot win against larger sellers. The sellers who hold up best in AI answers are the ones whose listings give the AI a reason beyond price: verified compatibility, a clear warranty, faster dispatch, a returns policy stated in one sentence, or a specification competitors omit. Price still matters — Adobe's own holiday analysis notes that price-conscious shoppers lean on AI tools to find deals — but price alone is a thin reason to be recommended.
The seven-step AI-readiness checklist
None of this requires rebuilding a store. It requires making the facts a machine can already find easier to find. Work through this list before Black Friday.
1. Write titles as answers, not slogans. Include the product type, the key attribute, and the use case: "Men's waterproof hiking boots, size 10, Vibram sole" beats "Amazing boots — best deal!" every time an AI tool is asked for waterproof hiking boots in size 10.
2. Put specifications in plain text. Materials, dimensions, weight, compatibility, battery life, and what's in the box should appear as readable text, not only inside images. AI tools cannot reliably read text baked into a photograph.
3. State price, stock, shipping, and returns on the product page. If the AI cannot confirm availability and delivery cost, it will recommend a listing that does. Keep sale pricing configured at the variant level so agents surface the real price.
4. Add a short FAQ to every important listing. Three or four genuine questions — "Does this fit X?", "How long does delivery take?", "What is the return window?" — written in full sentences. FAQ pages are among the highest-scoring content in Adobe's readability data, because they match the way people ask AI tools questions.
5. Publish a clear policies page. Return windows, shipping destinations, and delivery times, in plain language, on one page. Shopify's new knowledge-base feature exists for exactly this reason; on a marketplace, the same facts belong in the listing and shop policies.
6. Check your catalogue eligibility. On Shopify, open the catalogue view in the admin and fix anything that blocks inclusion. On a marketplace, check for listings flagged for missing item specifics or policy gaps. A blocked listing is invisible to agents no matter how good the product is.
7. Track AI referrals separately. Most analytics tools now show AI sources as a referrer. Watch them through November. If AI traffic is rising but not converting, the problem is usually on the landing page — missing facts, slow loading, or a checkout that demands too much before payment.
Three shifts worth thinking about
Beyond the checklist, three deeper changes are worth a seller's attention, because they will outlast this holiday season.
The first is the move from search results to answers. When ten links competed, being seventh still brought clicks. When one answer is given, being second brings nothing. That raises the stakes on every listing, but it also rewards depth: the seller who writes the clearest, most complete product information has a better chance of being the answer than the seller with the biggest ad budget.
The second is the move from storefront to fact sheet. The store still matters for checkout, trust, and service, but discovery increasingly happens elsewhere — inside a chat window, a voice answer, or an agent's shortlist. Investment follows the same logic: a pound spent on complete, accurate product data now reaches shoppers through every AI surface at once, while a pound spent on a homepage banner reaches only visitors who already arrived.
The third is the move from traffic to trust. Adobe's data shows AI shoppers spending longer on site and viewing more pages. They are verifying, not browsing. Reviews, clear policies, and consistent information across listings build the kind of trust that survives verification. Inconsistent prices, contradictory shipping claims, or missing specifications do the opposite — and an AI tool will surface the contradiction faster than any human comparison shopper.
None of this means human shoppers disappear. It means the most valuable shoppers increasingly arrive through a machine intermediary, and the sellers who win them are the ones whose information is built to be read by both.
The risks, stated honestly
A balanced guide should name the risks. AI tools sometimes present incorrect information with confidence, and a wrong recommendation can cost a seller a sale they never knew about or send a buyer a product that does not fit their need. Agents also compress choice: if only two or three products are quoted, the rest of the market is effectively hidden, which favours established sellers with complete data and makes it harder for a new seller to break in. And dependence cuts both ways — a seller who optimises entirely for one agent's format is exposed if that format changes. The sensible response is not to ignore AI shopping, but to keep the fundamentals strong: accurate data, honest descriptions, and customer service that turns a verified buyer into a repeat one.
What to do before Black Friday
Black Friday falls on 27 November 2026, and Adobe expects AI traffic to rise 95% that day alone. That leaves about four weeks. This week, fix titles, specifications, and policies on your ten best-selling listings. Next week, add FAQs and check catalogue eligibility. By mid-November, every listing you expect to sell over the holiday should state price, stock, delivery, and returns in plain text, and your analytics should be tracking AI referrals as their own channel. Sellers who do that work now will be the ones AI tools can find, understand, and recommend when the busiest shopping days arrive. The shoppers are already asking. The only question is whether your products are in the answer.
Frequently asked questions
What is an AI-referred shopper?
An AI-referred shopper is a visitor who arrives at a retail website by clicking a link inside an AI tool, such as a chatbot or AI search answer, after researching a product there. Adobe measures this traffic by clicks on links in AI platforms, and it has tracked the channel in the United States since October 2024.
Why do AI-referred shoppers convert better?
They usually arrive after the research and comparison stage is complete, so they are closer to a purchase decision than a casual browser. Adobe's September 2026 data shows they also spend 73% more time on site and add items to carts at a 53% higher rate, which points to verification rather than casual browsing.
What is Adobe's Citation Readability Score?
It is Adobe's measure of how much of a web page AI systems can understand and cite. The average US retail homepage scored 63% in September 2026, meaning more than a third of homepage content could not be read by AI systems. Pages that explain — specifications, guides, FAQs, and policies — score highest.
What is Shopify's shop.md?
Shop.md is a skill that turns an AI agent into a personal shopper connected to the buyer's Shop account. Reported on 19 October 2026 and included in Shopify's Fall '26 Edition on 28 October, it lets agents search more than one billion Shopify products, use the shopper's sizes and order history, check out with Shop Pay, and track orders inside the conversation.
Do I need to install shop.md on my store?
No. Shop.md is installed by the shopper into their own AI agent, not by the merchant. What merchants control is catalogue eligibility — complete product data, clear policies, and correct shipping settings — plus the knowledge-base facts agents receive about their store.
How can a small seller prepare for AI shopping before Black Friday?
Start with the listings that already sell: write titles as plain answers, put specifications in readable text, state price, stock, shipping, and returns clearly, add a short FAQ, and check catalogue or listing eligibility for gaps. Then track AI referrals as a separate channel through November to see what is working.
Conclusion
The headline number — 83% more revenue per visit — will get the attention, but the real story is quieter. Shopping is moving from a list of links to an answer, and the answer is built from facts. Sellers who give AI tools complete, honest, readable facts will be recommended. Sellers who do not will be skipped, however good their products are. Four weeks before Black Friday, that is the most practical opportunity in e-commerce: not a new advertising channel, but a better product page, written for the reader who now matters most — the machine that chooses what the customer sees.
For more guides on selling online, visit abdulrahimmughal.com.
Sources
- Adobe, "U.S. Holiday Shopping Season to Hit Record $275.1 Billion Online, Rising 6.7% YoY" (28 September 2026) — Adobe press release
- Retail TouchPoints, "Adobe: AI Traffic Driving Higher Revenue Per Visit for Retailers This Holiday Season" (29 October 2026) — Retail TouchPoints
- Search Engine Land, "AI traffic drives more revenue per visit for retailers: Adobe" (29 October 2026) — Search Engine Land
- Chain Store Age, "Adobe: AI e-commerce traffic soars in September" (29 October 2026) — Chain Store Age
- SMBTech, "AI-Referred Shoppers Now Worth Nearly Double Traditional Traffic For US Retailers, Adobe Finds" (30 October 2026) — SMBTech
- Digital Commerce 360, "Adobe projection: 2026 online holiday sales to reach $275.1 billion" (29 September 2026) — Digital Commerce 360
- E-Commerce News Australia, "Shopify unveils AI shopping tools in Fall '26 update" (29 October 2026) — E-Commerce News
- SMBTech, "Shopify Launches Personal Shopper AI Agents" (19 October 2026) — SMBTech
- Falia, "Shopify's Fall '26 Edition: AI agents become a sales channel to manage" (28 October 2026) — Falia
- Muhammad Umair Amjad, Threads post citing Adobe's 83% revenue-per-visit figure (29 October 2026) — Threads post
- @shoperatorai, Instagram reel on Shopify's shop.md and thin product data being skipped by AI agents (28 October 2026) — Instagram reel