Black Friday 2026 falls on November 27. For a growing share of shoppers, the first stop will not be your homepage, your app or your email. It will be an AI assistant that compares products, reads reviews and specs, and sends them to the one product page it thinks fits best.

In the last six weeks, three companies that see an enormous amount of real shopping data published numbers on exactly this: Adobe, which measures traffic to US retail sites, and Amazon and Shopify in their third-quarter results. Read together, they say something simple and uncomfortable: AI shoppers are now the most valuable visitors a retailer gets, and they decide before they arrive.

This article goes through what each of them reported, what the numbers do and don't prove, and what you can still fix in your catalog in the three weeks left.

AI shoppers now arrive ready to buy

Adobe has tracked traffic from AI services to US retail sites since October 2024. Its data covers more than one trillion visits, 100 million SKUs and 18 product categories, and it counts AI traffic only when a shopper clicks a link from an AI service through to a retailer.

Its latest report, published on October 29, shows how quickly the quality of that traffic has changed. In September 2026:

→ AI traffic converted 90% better than non-AI traffic. In August 2025 it still converted 9% worse.

→ AI traffic generated 83% more revenue per visit. In twelve months, an AI visit went from being worth less than half a typical visit to nearly twice as much.

→ AI visitors were more engaged. They spent 73% more time on site, viewed 25% more pages and added products to their cart at a 53% higher rate.

The volume is growing too. AI traffic to US retail sites rose 71% year over year in September, and Adobe forecasts a 130% increase over the whole holiday season, with the peaks on the days that matter most: +159% on Thanksgiving, +95% on Black Friday and +88% on Cyber Monday. Adobe also expects the revenue-per-visit advantage of AI traffic to double during the season.

The reason is not mysterious. A shopper who asks an assistant for the best waterproof hiking boots under 150 euros has already compared, filtered and shortlisted before landing on a site. The research happened somewhere else. By the time the visit arrives, it is a buying visit.

Amazon and Shopify: the agents are already selling

The two largest commerce platforms in the West reported the same shift from the inside.

Amazon. On its Q3 call on October 29, Amazon said that more than 400 million customers used Alexa for Shopping, its agentic shopping assistant, in the last twelve months, up from 350 million at the end of July, according to Modern Retail. Customers who use it spend 50% more than those who don't. Amazon added more than 100 capabilities before the holidays, from gift lists and registries to managing Subscribe and Save, all by conversation. Andy Jassy was explicit about the ambition: "we believe Alexa for Shopping is going to be the primary shopping agent customers use to shop Amazon."

We wrote in September that Amazon's Q2 ended the debate on whether agentic commerce converts. Q3 shows it scaling.

Shopify. On November 3, Shopify reported that AI traffic to its merchants' stores grew 2.6 times year over year, and that orders grew even faster, 3.3 times, as reported in coverage of the earnings call. Three details from the call matter more than the multiples:

→ AI sends shoppers straight to product pages, at about three times the rate of organic search, and those shoppers convert 64% more often.

→ 74% of AI-attributed purchases came from outside Shopify's top 100 product categories. Agents find long-tail products that search and ads rarely surface.

→ Q3 was the first quarter of delegated purchasing, with personal agents completing a purchase on the shopper's behalf.

What the numbers have in common: the product page is the new landing page

Put the three sources side by side and the same pattern appears.

AI shoppers skip the homepage. They land on a specific product, often one you never promoted, with their decision mostly made. Whether they arrive at all depends on whether the assistant understood that product well enough to recommend it.

That understanding comes from the product data, not from your campaign. Adobe measured how much of a page AI systems can actually read and cite: the average US retail homepage scored 63%, so more than a third of its content is invisible to them. The best-scoring category, electronics at 79%, is the one full of specs, setup guides and troubleshooting. Adobe's own conclusion is that the content AI reads best is content that explains: product details, buying guides, FAQs and return policies.

This is where Black Friday gets interesting. Most holiday preparation goes into things an agent never reads: the countdown banner, the discount badge, the landing page hero. An agent compares attributes, prices, availability and answers to the shopper's question. If your product page doesn't state them, the agent recommends a product that does.

The long-tail finding from Shopify makes the point sharper. When three quarters of AI purchases come from outside the top categories, optimizing only your fifty best sellers is not enough. The agent will look at the whole catalog, and every product with thin or inconsistent data is a product it can't confidently recommend.

Read the numbers with care

These figures are strong, but they need context before they drive decisions.

→ Adobe's holiday figures are forecasts, not results. Only the September data is measured.

→ Adobe counts clicks. A purchase completed by an agent without a visit to your site is not in its AI traffic, and an AI answer that shapes a decision made later through search shows up as search. The real influence of AI is larger than the referral numbers, and harder to see.

→ Growth is measured from a small base. Adobe's releases don't give AI's share of total retail traffic, and a 130% increase on a small channel is still a smaller channel than search.

→ Amazon and Shopify report their own metrics, on calls designed to show momentum to investors, and the Shopify figures come from statements on the call, not from the written results.

None of this weakens the direction. It means the right reading is not "AI will be your biggest channel this Black Friday" but "AI already brings your highest-intent shoppers, and it chooses which of your products they see".

A three-week catalog checklist

Three weeks is not enough to rebuild a catalog. It is enough to fix the things AI systems read first, on the products that matter for the season. Changes also need time to be crawled and reflected in feeds and AI answers, which is why the order below matters.

Week 1 (November 9 to 15): make your products readable

→ Pick the products you want agents to recommend: your seasonal best sellers and gift products, plus a sample of long-tail items, since that is where AI purchases go.

→ Complete the attributes shoppers filter on: material, fit and size guidance, dimensions, compatibility, who the product is for. State them on the page as clear specs, not only inside marketing prose.

→ Check that AI crawlers can reach your product pages. Look at robots.txt and at your CDN or firewall settings for OAI-SearchBot, PerplexityBot, Claude-SearchBot and Bingbot. Many stores block them without knowing it.

Week 2 (November 16 to 22): answer the questions behind the purchase

→ Add the answers shoppers ask an assistant before buying: what it is for, how it fits, how it compares with your other models, how to care for it, whether it works as a gift. A short FAQ on the product page often covers most of them.

→ Make offers machine-readable. The Black Friday price, sale dates, availability, shipping times and return policy must be the same on the page, in the structured data and in your product feed. An agent that sees two prices trusts neither.

→ Refresh your feeds, including the conversational attributes Google added to Merchant Center, and the product feed if you sell through ChatGPT.

Week 3 (November 23 to 27): freeze, measure and watch

→ Freeze content changes on the key products, and keep only price and stock updates flowing.

→ Measure AI traffic properly. In GA4, group the AI sources (ChatGPT, Perplexity, Gemini, Copilot, Claude) in their own channel, so you can compare their conversion and revenue per visit with the rest during the peak.

→ Ask the assistants yourself. Run the ten questions your customers would ask on ChatGPT, Gemini and Perplexity, note which products they recommend and why, and repeat after the peak.

If you want to check where your product pages stand, our free AI Readiness audit scores your product pages on the attributes, questions and data that AI agents read, and our product page checklist lists the 46 checks we run.

The banner doesn't reach the agent

The first AI-powered holiday season was a forecast in July. The numbers from Adobe, Amazon and Shopify say it has arrived, and that it rewards a different kind of preparation.

Budget can move overnight in November. Product data can't. This Black Friday, the shoppers most likely to buy will meet your products through an agent first, and the agent only knows what your product pages actually say.

Sources

→ Adobe, U.S. Holiday Shopping Season to Hit Record $275.1 Billion Online, September 28, 2026.
→ Adobe, AI traffic now generates 83% higher revenue per visit, October 29, 2026.
→ Amazon, Q3 2026 earnings report, October 29, 2026, and Modern Retail, Amazon says more than 400M customers have used Alexa for Shopping, October 29, 2026.
→ Shopify, Q3 2026 financial results, November 3, 2026, and Deployed (AIM Media House), AI orders on Shopify grew 3.3x, November 4, 2026.