Online stores
Shoppers ask AI, then buy on your site: the 2026 holiday numbers.
Adobe expects 130% more clicks from AI assistants to US retail sites this holiday season than last, and the surveys say few shoppers will let an assistant do the buying. In Deloitte's 2026 holiday survey, 39% of US shoppers said generative AI is their primary tool for at least one holiday shopping task, and 16% said they would let it complete a purchase. The assistant helps shoppers choose, most still buy for themselves, and the clicks Adobe counts land on retailers' own sites. Below are the figures that matter to an online store, each with its publisher and date, and then what to do and when.
Key takeaways
- AI is a research tool this season, not a checkout. 39% of US shoppers call generative AI their primary tool for at least one holiday shopping task, and 16% would let it complete a purchase (Deloitte, October 2026, 4,000 consumers).
- The clicks it sends are growing fast, from a small base. Adobe forecasts AI traffic to US retail sites up 130% on 2025, after a 693.4% rise last season that Adobe itself said came from a modest base of users. It gives no share of total traffic.
- Much of the shortlist forms inside the answer. Of US shoppers who use AI, 43% used it to narrow a shortlist and 37% made a final purchase decision based on its recommendation (Menlo Ventures, September 2026).
- Shoppers who use AI consult more sources, not fewer. Gartner found they consult 4.48 sources for holiday ideas against 3.24 for those who don't (October 2026). In an earlier Gartner survey, 54% of AI shoppers said they had to double-check everything the tools told them. Your product page needs to answer what they check.
What the holiday surveys found
39% of US shoppers say generative AI is their primary tool for at least one holiday shopping task, and only 16% say they would let it complete a purchase. The 2026 Deloitte Holiday Retail Survey, published on 15 October, asked 4,000 US consumers between 26 August and 10 September, with a margin of error of plus or minus 2 points.
Adobe forecasts AI traffic to US retail sites up 130% on last year between 1 November and 31 December (Adobe, 28 September 2026). Adobe sells analytics software to retailers and forecasts from its own data. AI traffic is Adobe's count of shoppers clicking a link in an AI service through to a retail site. Of the three days it named, Adobe expects the largest rise on Thanksgiving, then Black Friday and Cyber Monday. In the same release, an Adobe survey of 5,000 consumers in July found that 77% of those who have used AI for online shopping said they feel more confident in their purchase. In the same survey, 69% said they are less likely to return an item they bought.
Year-on-year growth, not a share of traffic: Adobe does not say what part of all retail visits will come from AI. Adobe forecast, 28 September 2026.
Of US shoppers who use AI, 50% used it to compare prices, 43% to narrow a shortlist, 37% made a final purchase decision based on its recommendation and 26% let it complete a purchase. The figures come from Menlo Ventures' 2026 State of Consumer AI report (16 September 2026; 5,067 US adults surveyed with Morning Consult in July). Menlo is a venture firm that invests in AI companies, Anthropic among them, so read it as a large survey from an interested party.
Each bar is a share of US shoppers who use AI, not of all shoppers, and the widest is 50%. Menlo Ventures with Morning Consult, 5,067 US adults, July 2026.
34% of US consumers are very or extremely willing to use generative AI tools or chatbots for holiday gift recommendations, Gartner reported on 7 October 2026, from a survey of 1,004 US consumers in August. Those who already use AI for holiday shopping inspiration consult an average of 4.48 sources, against 3.24 for those who do not. Retailer websites, ordinary web searches and wishlists from friends and family remain the most common sources of ideas.
Three numbers on letting AI buy
Three surveys published in 2026 measured how many shoppers would let AI decide or buy, or already have, and came back with 11%, 16% and 26%. Those are not three readings of one number. Each asks a different question of a different group:
| Figure | What was measured | Who was asked | Source |
|---|---|---|---|
| 11% | The highest willingness to let AI make the purchase decision, reached in lower-stakes categories such as personal care and household supplies | US consumers | Gartner, 27 May 2026 |
| 16% | Would let generative AI complete a purchase | 4,000 US consumers, August to September 2026 | Deloitte, 15 October 2026 |
| 26% | Let AI complete a purchase in 2026 | US shoppers who already use AI | Menlo Ventures, 16 September 2026 |
Menlo's figure is the highest partly because it counts only shoppers who already use AI. Even among them, about three in four did not let it buy. On all three measures, most people do not hand the purchase to an AI.
The traffic figures need the same care. Adobe measured AI traffic to US retail sites up 693.4% in the 2025 season and wrote in the same release that "the base of users remains modest" (Adobe, 7 January 2026). A 130% rise on that base is fast growth in a channel that is still a minority of visits, and Adobe publishes no share of total traffic to say how small a minority. Last season, Adobe saw AI services used most in categories including video games, toys, appliances, electronics and personal care products.
What it means for an online store
Much of the choosing happens before the visit
If 43% of shoppers who use AI narrowed a shortlist with it and 37% made the final decision on its recommendation, then for those shoppers much of the choosing happens before anyone opens a product page. The click Adobe counts goes to a store the answer named. A store missing from the answer misses the clicks that answer sends. Which tools show whether AI assistants recommend your products sets out what each kind of tool can see of that answer, from ChatGPT to Amazon's own assistant.
The visit is a check
Gartner's two findings point the same way. Shoppers who use AI for ideas consult more sources than those who do not. And in a Gartner survey of 846 US consumers (November to December 2025), more than half of those who had used AI while shopping said they had to double-check everything it told them (Gartner, 27 May 2026). So the visitor an assistant sends often arrives to confirm something: the price, whether it is in stock, whether it arrives before the holiday, what the returns are. Salesforce, which sells commerce software to retailers, reported that in the 2025 season, shoppers sent to retailer websites by AI search channels converted nine times as often as shoppers from social media. That comes from its data on 1.5 billion shoppers worldwide (Salesforce, 8 January 2026). A product page that answers those questions in plain words gives that visitor what they came to confirm.
Amazon's assistant is its own channel
Alexa for Shopping is Amazon's AI shopping assistant, formerly Rufus, now combined with Alexa+. Amazon said in its second-quarter 2026 results that the assistant's active users worldwide were close to doubling, with interactions up more than five times year on year (Amazon, 30 July 2026). US customers who use it spend on average over 40% more per order than those who do not. That compares customers who chose to use the assistant with those who did not, so it cannot say the assistant caused the bigger orders. Amazon's figures cover its own store: a store that also sells there competes in a different answer from the one ChatGPT, Gemini or Google's AI Mode, its chat-style search, gives. Our comparison of AI visibility tools for online stores lists which tools say they track products in ChatGPT Shopping, AI Mode and Alexa for Shopping.
What to do, and when
- This week: make sure the assistants can read your store. Every figure above assumes an assistant can fetch your product pages. Our free crawler check reads your robots.txt and reports sixteen AI crawlers, including OpenAI's three separate agents. If your host or CDN has a bot-blocking setting, check that too.
- This week: ask the assistants what your buyers ask, three times each. Gift questions with a budget, such as a present for a runner under $50; where-to-buy questions with a deadline, such as which store can deliver before Christmas; and your store against the one you lose sales to. Write down which stores were named and which pages were cited. That list, not a general ranking, is the shortlist you are trying to join.
- Before Thanksgiving: put the answers on the product page, in words. Price, stock, the last order date for holiday delivery and the returns window, written on the page rather than inside an image or a pop-up, and the same values in your Product structured data. The JSON-LD schema guide has the Product markup, the code that describes a product to search engines, written out in full.
- Before Thanksgiving: answer the shortlist question on your own site. A gift guide or comparison page that says plainly which product suits whom gives an assistant a passage it can use when it narrows the choice. Keep it current: a guide that recommends something out of stock in December sends the shopper to a product they cannot buy.
- Through December: check every week, not once. Answers move on their own, and we measured how much across 391 answers about AI visibility tools. Asked the same question at the next run, one to fifteen days later, the engines named 54% of the same tools again, and only 5% of repeat answers kept the whole list (1 to 25 September 2026). The questions were about software, not gifts, so read this as a sign of movement rather than a figure for your category. The weekly routine sets out how to check, free or with a tracker.
Do not:
- Quote Adobe's 130% as a share of your sales. It is growth on a base Adobe does not publish.
- Rebuild checkout for AI agents before the product page is right. No survey here puts the share who would let AI buy, or already have, above about one in four, and that highest figure counts only shoppers who already use AI.
- Block AI search crawlers to keep your prices away from them. An assistant that cannot read a page is less likely to cite or link it.
The bigger picture: forecasts for 2030
Bain forecasts US agentic commerce at $300 to $500 billion by 2030, roughly 15 to 25% of e-commerce (Bain, 17 December 2025). The definition matters more than the number. Agentic commerce is Bain's term for purchases that third-party or retailer-hosted AI agents initiate, influence or complete, not journeys that only use AI to search or discover. In the same note Bain put the share of US consumers already using generative AI for product research and comparison at 30 to 45%. McKinsey, in a report with ICSC, estimates up to $1 trillion of US business-to-consumer retail revenue from agentic commerce by 2030 (ICSC, 27 April 2026). The ICSC page does not say how McKinsey defines the term.
Both forecast 2030, on different definitions, and neither describes this season. This season's surveys describe something nearer: an assistant that researches, narrows and then sends the shopper to a store. The work that pays this December is pages an assistant can read and a product page that answers what shoppers check. An agent buying on a shopper's behalf would need the same pages, so none of it is wasted if the forecasts come true.
Where Greater Than Services fits
Our own product, so read this as the vendor's description, not part of the measurement above.
Greater Than Services is a workspace per brand that puts your buyer questions to five AI engines every week, records which answers named you and what they cited instead, shows your Google rank beside each answer, and drafts the fixes.
- If you need to know whether assistants name your store this season, it asks your buyer questions every week and records which answers named you and which stores came up instead.
- If you need to see what the shortlist was built from, it lists the sources each answer cited.
- If you need a page that answers what shoppers ask, the GEO Writer, the workspace's page-drafting tool, drafts it and can send it to Shopify or WordPress as an unpublished draft, so nothing goes live until you publish it.
- If you need Google as well, your Google rank for the same searches sits beside each answer, and Google AI Overviews is read off the live results page for your country.
It tracks whether answers name your store. It does not track individual products, AI shopping cards or Alexa for Shopping; the comparison of tools for online stores linked above lists the ones that say they do.
Plans start at $49 a month, billed yearly, for one brand on ChatGPT and Google AI Overviews; Growth and Agency cover all five engines. The three-day trial is on one brand and needs a card.
Common questions
How many shoppers will use AI for holiday shopping in 2026?
It depends on the measure, and the surveys should not be averaged. In Deloitte's 2026 Holiday Retail Survey of 4,000 US consumers, 39% said generative AI is their primary tool for at least one holiday shopping task. Gartner found 34% of 1,004 US consumers very or extremely willing to use generative AI tools or chatbots for holiday gift recommendations. Bain put the share of US consumers using generative AI for product research and comparison at 30 to 45% in December 2025.
Will AI assistants buy things for shoppers this holiday season?
Mostly not. Deloitte found 16% of US shoppers would let generative AI complete a purchase. Gartner found that willingness to let AI make the purchase decision topped out at 11%, and that was in lower-stakes categories such as personal care and household supplies. Menlo Ventures, which invests in AI companies, found that 26% of US shoppers who use AI let it complete a purchase in 2026. Of those shoppers, 50% compared prices and 43% narrowed a shortlist with it; about three in four have not let it buy.
How much traffic will AI send to online stores this holiday season?
Adobe forecasts AI traffic to US retail sites up 130% on 2025 between 1 November and 31 December. It counts that traffic as shoppers clicking a link through to a retail site, and of the three days it named, expects the largest rise on Thanksgiving, at 159%. That is a growth rate, not a share. Adobe does not say what part of all retail visits will come from AI, and it described last season's 693.4% rise as coming from a modest base of users.
Do shoppers who come from AI assistants buy?
Two companies that sell to retailers report encouraging figures. Salesforce reported that in the 2025 holiday season, shoppers sent to retailer websites by AI search channels converted nine times as often as shoppers from social media referrals, in its data on 1.5 billion shoppers worldwide. In Adobe's July 2026 survey of 5,000 consumers, 77% of those who have used AI for online shopping said they feel more confident in their purchase, and 69% said they are less likely to return an item they bought.
Sources
Every external figure was fact-checked against its publisher's page, press release or filing on 18 October 2026. Each survey is given with its sample and date in the text above, and vendor studies are named as such where they appear.
- Deloitte, 2026 Deloitte Holiday Retail Survey, 15 October 2026
- Adobe, 2026 holiday season forecast, 28 September 2026, and 2025 holiday season results, 7 January 2026
- Menlo Ventures, 2026: The State of Consumer AI, 16 September 2026
- Gartner, holiday gift recommendations survey, 7 October 2026, and AI shopping help survey, 27 May 2026
- Amazon, second-quarter 2026 results, SEC filing, 30 July 2026
- Salesforce, 2025 holiday shopping data, 8 January 2026
- Bain & Company, 2030 agentic commerce forecast, 17 December 2025
- ICSC with McKinsey & Company, Shopping in the age of AI, 27 April 2026
The repeat rate is our own, counted from the answers behind our AI visibility tools ranking, which sets out the questions, the engines and the dates and publishes the data under CC BY. If a figure here does not match its source, tell us and we will correct it.