AI Shopping Statistics: What the Data Shows for Retailers

Checked 19 September 2026. This article is reviewed quarterly, because AI referral, conversion and visibility data is changing faster than anything else we publish.

Key Takeaways

• Adobe Digital Insights analyses more than a trillion visits to US retail sites and over 100 million SKUs, which makes it the most detailed public source on AI shopping (Adobe).
• AI traffic to US retail grew 393% year on year in Q1 2026, slowing to 138% in May and 62% in July 2026 as the comparison base grew.
• The widely quoted 1,151% December 2025 figure is not year on year growth. It measures growth in AI visit share against Adobe’s October 2024 baseline, so it should never be charted alongside the figures above.
• The quality story runs the other way. AI referred visitors converted 38% worse than other traffic in March 2025 and 60% better by July 2026.
• In May 2026 they generated 53% more revenue per visit, spent 53% longer on site, viewed 23% more pages and were 36% less likely to bounce.
• Adobe’s April 2026 benchmark scored US retail product pages at 66% AI readability, the worst of any page type it measured (Adobe).
• Shopify reported that in Q1 2026 AI driven traffic to its stores grew eight times year on year, with orders from AI powered searches up nearly 13 times.
• UK evidence exists but is thinner: Adobe reported UK retail AI traffic up more than 1,200% between August 2024 and June 2025, and 329% year on year over the 2025 holidays.

Quick Answer

AI shopping traffic is still small next to established channels, but its growth and its quality are now hard to ignore. Adobe’s US retail data shows year on year growth normalising, from several hundred percent in late 2025 to 62% by July 2026, while conversion moved from 38% worse than other traffic to 60% better in the same period. Shopify reports the same direction on its own platform, with AI driven traffic up eight times year on year in Q1 2026. The practical problem for merchants is visibility rather than volume, because Adobe scored the average retail product page at 66% machine readable. If an assistant cannot read your specifications, it cannot recommend your product.

Table of Contents

• How to Read AI Shopping Statistics
• How Much AI Traffic Is There?
• Does AI Traffic Convert?
• Why AI Traffic Behaves Differently
• How People Actually Use AI to Shop
• Is Your Content Even Readable by AI?
• What Changes for Shopify Merchants
• What This Means for SEO
• What We Would Do About It
• How to Measure AI Traffic in Your Own Store
• What the Numbers Mean
• Methodology and Update Schedule
• Frequently Asked Questions
• Why Matters Perspective
• Final Thoughts
• Related Articles

AI shopping has moved from an interesting experiment to a measurable channel, and the reversal in traffic quality matters more than the enormous growth percentages that make headlines. There are good reasons for caution, though. Most of the detailed data comes from one provider, covers US retail, and reports percentage change from a small base. This guide keeps those figures separate rather than blending them, and concentrates on what the evidence actually supports and what merchants should change if more buying decisions start with an assistant.

How to Read AI Shopping Statistics

AI ecommerce statistics need more qualification than most marketing benchmarks, for four reasons.

1. Much of the best data comes from one provider. Adobe Digital Insights is the most detailed public source we have found, analysing over a trillion US retail visits, but Adobe customers are not a representative sample of every retailer, country or platform. Useful evidence, not a census.
2. Most of the detail is US retail. Conversion, revenue per visit, engagement and readability figures predominantly cover US sites. Adobe has published UK evidence, including AI traffic to UK retailers growing more than 1,200% between August 2024 and June 2025 and referrals up 329% year on year over the 2025 holidays, but we have found no equivalent 2026 UK dataset with the same depth, so we do not relabel US benchmarks as UK statistics.
3. Huge percentages start from a tiny base. A channel can grow 1,000% and still be a small share of total visits. Adobe has repeatedly described AI traffic as modest against paid search and email, and does not publish a clean absolute share, so we will not invent one.
4. Not every percentage uses the same baseline. This is the easiest mistake to make with this data, and it is worth its own explanation below.

The Baseline Trap

You will see 1,151% for December 2025, 393% for Q1 2026, 138% for May and 62% for July, and it is tempting to chart all four as a collapse in growth. They do not share a denominator. The 393%, 138% and 62% figures are year on year growth rates. The 1,151% figure describes growth in AI driven visit share relative to Adobe’s October 2024 tracking baseline. There is a genuine deceleration in year on year growth, but it has to be shown using comparable numbers.

How Much AI Traffic Is There?

The interesting story in 2026 is not growth. It is growth beginning to normalise.

Year on year growth has moved from several hundred percent towards double digits, which does not mean AI shopping is shrinking. It means the comparison base is much larger than it was. Against Adobe’s October 2024 starting point, the separate baseline series reported 1,151% in December 2025, 1,324% in May 2026 and 1,219% in July 2026. Read together, the two series say something sensible: the explosive phase is slowing, and AI referral traffic remains many times larger than when measurement began. Neither "AI is growing 1,000%, search is dead" nor "growth fell to 62%, AI shopping is over" follows from the evidence.

Does AI Traffic Convert?

This is where the data becomes commercially interesting, because AI referral traffic initially converted badly and no longer does.

The supporting metrics moved with it. In March 2026, AI referred visitors generated 37% higher revenue per visit, spent 48% longer on site, viewed 13% more pages, bounced 32% less and were 12% more engaged. By May 2026 those gaps had widened to 53% more revenue per visit, 53% longer on site, 23% more pages, 36% lower bounce and 15% higher engagement. By July 2026, Adobe reported 14% higher engagement, 59% more time on site, 33% less bouncing and a 28% higher rate of adding products to the basket.

Why AI Traffic Behaves Differently

The explanation is in the sequence. An AI assistant is increasingly the research layer immediately before the visit, so people arrive having already narrowed the options, checked the specifications and formed an intention. That produces a visitor who needs confirmation rather than education, which is why the landing page matters more than it used to and why generic category pages disappoint this traffic. Adobe’s survey supports it: 79% of AI shopping users said it made them more confident in a purchase, and 69% said they were less likely to return an item bought with AI help.

How People Actually Use AI to Shop

Adobe’s March 2026 survey of more than 5,000 US consumers found 39% had used AI assistants for online shopping, and 85% of those said it improved the experience. Two thirds, 66%, said they trusted the tools to provide accurate results, 80% said they use assistants more than they used to, 50% click the links the assistant provides and 27% have bought directly through them.

An earlier Adobe survey of 5,000 US consumers found the specific jobs people give AI: 53% product research, 40% recommendations, 36% finding deals, 30% shopping lists, 30% gift ideas, 29% discovering unique products and 26% virtual try on. Two findings matter commercially. Among people who had used AI for shopping, 73% called it their primary source of product research, and 83% said they were more likely to use it for larger or more complex purchases. That is exactly where detailed product information decides the sale.

Is Your Content Even Readable by AI?

This is where AI statistics become a website problem. A merchant can have excellent products and still give AI systems very little usable information about them. Adobe’s April 2026 analysis used its AI Content Visibility Checker to score how much retail content machines could read.

There is a correction here for an assumption merchants often make. Product pages were not the best performing pages in the benchmark. They were the worst, at 66%, which means roughly a third of product page content was not available in a form the checker considered machine readable. Adobe notes that a 50% score means half of a page is unreadable to machines, so 66% is not a pass mark.

By July 2026 Adobe widened the analysis to a broader group of US retailers and average homepage readability fell to 61%, implying nearly 40% of homepage content was not fully machine readable, with sector scores ranging from apparel at 76% down through electronics at 70%, cosmetics 68%, sporting goods 67%, furniture and home 64%, general merchandise 63% and grocery at 59%. The April and July figures use different samples, so they are not a clean time series, but the message holds: a site can look excellent to a person while hiding important information from a machine.

What AI Needs to Understand About a Product

For ecommerce that usually means ingredients, materials, dimensions, weight, capacity, compatibility, sizes, colours, variants, certifications, care instructions, subscription options, delivery, returns, warranty, stock status and price. Putting any of that only inside an image, an interactive component or a beautifully designed graphic hides it. Humans may see it. The system recommending products may not. Adobe also found that retailers with high AI visit share had a 52% readability advantage on homepages, 32% on search results pages, 23% to 30% on content and buying guides and around 14% to 15% on product detail and brand pages, with fewer content gaps associated with higher readability throughout.

What Changes for Shopify Merchants

Shopify is not treating this as a theoretical future channel. It reported that in Q1 2026 AI driven traffic to Shopify stores grew eight times year on year, orders from AI powered searches grew nearly 13 times, and new buyers were ordering through AI channels at nearly twice the rate of other channels. Its own analysis also found more than half of AI referred sessions started on product detail pages, against about 20% for organic search, that those product page sessions converted at nearly 50% higher rates than organic search, and that AI referred orders carried 14% higher average order values. That is a different dataset from Adobe’s, so the numbers should not be combined, but they point the same way.

• ChatGPT. Shopify says product discovery is live for millions of merchants selling to US buyers, wherever the merchant is based, with checkout completed on the merchant’s own Shopify checkout in an in app browser or a separate tab.
• Agentic Storefronts. The sales channel distributing product information to supported AI surfaces including ChatGPT, Google AI Mode and Gemini, Microsoft Copilot and Meta. It is active by default for eligible stores, and controlled under Sales channels, then Agentic, in admin.
• Shopify Catalog. Structures, cleans and syndicates product data to connected AI channels, sharing titles, descriptions, images, pricing and availability. That makes something very ordinary strategic: the quality of your product data.
• Universal Commerce Protocol. Announced on 11 January 2026 and co developed with Google, an open standard for commerce between AI agents and merchants covering discovery, carts and checkout.

For the buying channel itself, our agentic commerce guide covers how the transactions work. This article stays on the statistics, the visibility problem and what merchants should change on their own sites.

What This Means for SEO

AI shopping does not end conventional SEO. It changes the job. Rankings, keywords, links and clicks still matter, and a second question joins them: can a system confidently extract an answer from your content and recommend your product? That rewards things good ecommerce SEO should already be doing, including descriptive product names, clear specifications, useful category copy, comparison information, complete descriptions, FAQs based on real buying questions, clear delivery and returns information, and accurate structured data. Our Shopify SEO guide covers the conventional side.

On whether AI search reduces ecommerce traffic, the honest answer is that it can reduce some informational clicks, because an assistant can answer without sending anyone anywhere. The retail data so far does not show AI eliminating merchant traffic. It shows a new referral channel growing quickly. The likely change is that part of discovery and research moves off your website, and the clicks that remain arrive later in the decision, which makes the destination page more important rather than less.

What We Would Do About It

If we were auditing a Shopify store for AI visibility today, we would work in this order.

1. Make the facts explicit. For skincare that means ingredients, skin type, fragrance, allergens, volume and certifications. For electronics, dimensions, connection type, compatibility, power and warranty. For pet food, ingredients, animal size, age, feeding guidance and pack weight. Do not make a system infer the product from lifestyle photography.
2. Put important information in text. If a critical specification exists only inside an image, carousel or interactive element, add a readable version. That helps accessibility at the same time.
3. Answer buying questions on the page. Is this suitable for sensitive skin, will it fit this model, how long does a pack last, can I pause the subscription, does it contain nuts. Answer them where the decision happens, which our product page design guide covers from the human side.
4. Use structured data properly. Product, offer, price, availability and review data should corroborate what is visible on the page. It cannot compensate for a page that says almost nothing.
5. Check what machines can actually see, rather than assuming rendered content is accessible because it looks fine in a browser. Test descriptions, specifications, pricing, variants, FAQs, delivery and returns.
6. Fix the technical foundation. Across 265 Shopify stores audited by Why Matters, 196 were still on legacy themes, roughly 74%, and stores we moved to Online Store 2.0 gained an average 23 points of mobile PageSpeed (our speed research). AI readability work belongs alongside speed, theme architecture, product information and accessibility, not on top of an ageing foundation.
7. Measure citations as well as clicks. Traffic tells you what happened after an assistant sent somebody to you. It does not tell you whether you were considered and rejected. Track prompts such as best category for a need, brand alternatives, products containing an ingredient, compatibility questions and brand versus competitor, then record whether you appear, whether you are cited, which page is cited and whether the information is correct.

How to Measure AI Traffic in Your Own Store

Start in Google Analytics, which added a dedicated AI Assistant channel to its default channel group on 13 May 2026, automatically assigning an AI assistant medium when a referrer matches a recognised assistant, including ChatGPT, Gemini, Deepseek, Copilot and Grok. One detail catches people out: traffic from Google AI Overviews and AI Mode is classified as organic search rather than AI Assistant, so your AI numbers are not the whole picture. Google also documents how to build a custom AI assistant channel using referrers for services including ChatGPT, Gemini, Copilot, Claude and Perplexity.

Then compare more than sessions. Look at conversion rate, revenue per visit, average order value, pages per visit, bounce and which landing pages AI traffic arrives on, using our conversion rate guide for the measurement basics. Shopify merchants should also check Agentic Storefront reporting, which can show sales, orders, sessions and conversion for individual AI channels, with orders attributed by channel or referrer in admin. Inspect the landing pages themselves, because if more than half of AI sessions start on product pages, those pages are carrying the channel.

What the Numbers Mean

1. The growth rate is slowing, which is normal as a base grows and is not evidence that AI shopping is fading.
2. Traffic quality is moving the other way. A channel that converted 38% worse now converts 60% better, which is an unusual reversal in any channel.
3. AI is becoming part of research before the visit, so people arrive further down the decision than they used to.
4. Product information is infrastructure. With product pages scoring worst on readability, the gap between what a site shows a person and what it gives a machine is a commercial problem.
5. Your own data matters more than Adobe’s. Their benchmarks tell you the direction. Your analytics tell you whether it is happening to you, which is also the argument in our UK Shopify statistics article.

Methodology and Update Schedule

This article was checked on 19 September 2026. Traffic, conversion, engagement and readability figures come from Adobe Digital Insights reports published between January and August 2026, covering US retail unless stated otherwise, with UK figures labelled as UK. We keep the year on year series and the growth from baseline series separate throughout, because they use different denominators and blending them produces a misleading chart. Where Adobe does not publish an absolute share of all retail traffic, we do not estimate one.

Shopify figures come from Shopify’s own published data and help documentation, and are not combined with Adobe’s because the datasets differ. The 265 store audit findings are Why Matters data rather than platform wide statistics. On updates, we review this article quarterly, because Adobe publishes quarterly and these numbers move faster than anything else on this blog.

Frequently Asked Questions

Does AI traffic convert?

Yes, and increasingly better than other channels. Adobe found AI referred US retail traffic converted 38% worse than other traffic in March 2025, then 42% better in March 2026, 54% better in May and 60% better in July 2026.

How do I get my products into AI search?

Make the facts about each product explicit in text, answer real buying questions on the page, keep structured product data accurate, and make sure what a machine can read matches what a shopper sees. Shopify merchants should also check their Agentic Storefronts and Catalog settings.

Will AI search reduce ecommerce traffic?

It can reduce some informational clicks, because an assistant can answer without sending anyone anywhere. The retail data so far shows a new referral channel growing rather than merchant traffic disappearing, with visitors arriving later in the decision.

How do I track ChatGPT traffic?

Google Analytics added an AI Assistant channel to its default channel group in May 2026, assigning an AI assistant medium to recognised assistants including ChatGPT, Gemini, Copilot and Grok. Traffic from Google AI Overviews and AI Mode is still classed as organic search.

Is AI traffic better than Google traffic?

On Adobe's US retail data it converts better and generates more revenue per visit, but it is a much smaller channel. Treat it as a high quality channel worth preparing for rather than a replacement for search.

How much ecommerce traffic comes from AI?

Adobe publishes growth rates rather than a clean absolute share, and describes AI traffic as still modest against established channels such as paid search and email. Any specific percentage of all ecommerce traffic should be treated sceptically.

What is AI readability?

It measures how much of a page a machine can actually extract. Adobe's April 2026 benchmark scored US retail product pages at 66% on average, meaning roughly a third of product page content was not available in a form its checker considered machine readable.

Is AI shopping growing or slowing down?

Both, depending on the measure. Year on year growth in AI traffic to US retail fell from 693% over the 2025 holidays to 62% in July 2026, because the comparison base is much larger, while traffic quality improved throughout.

Why Matters Perspective

The temptation with AI statistics is to pick the biggest number and build a strategy on it. We would do the opposite. The two figures that should change what a merchant does this quarter are the conversion reversal and the 66% product page readability score. The first says this traffic is worth having. The second says most retailers are not ready to receive it.

What makes that encouraging is that the work is not exotic. Explicit specifications, buying questions answered on the page, accurate structured data, a fast modern theme and information that exists as text rather than only as design are all things that improve the experience for customers first. If AI discovery turns out to matter less than expected, none of that effort is wasted. If it matters more, the stores that did it will be the ones being recommended.

Why Matters is a Shopify Select partner with Verified Skills across development and marketing. Our pricing is published, retainer clients are billed one month in arrears and never tied into long contracts, and every development project carries a 3 month guarantee. See our Shopify packages or email us to talk through your store.

Final Thoughts

AI shopping statistics are easy to sensationalise and easy to dismiss. The measured version is more useful. Growth is normalising from a small base, the traffic that arrives is now the best converting in the dataset, people are using assistants for research before they reach your site, and most retail product pages give those assistants less to work with than they give a human. That last point is the one you control. Make the facts readable, answer the questions buyers actually ask, keep the technical foundation sound, and measure what reaches you rather than trusting anybody else’s benchmark, including this one.

Related Articles

• Agentic Commerce for Shopify Stores: A Practical UK Guide
• Shopify Product Page Design: What Actually Converts
• How to Optimise Your Shopify Store for SEO
• Shopify Speed Score: How to Measure and Improve It
• Shopify Statistics UK: Store Numbers, Market Share and Growth
• Shopify App Statistics and What They Mean for Merchants
• Why Your Shopify Conversion Rate Matters and How to Measure It
• 20 of the Best Shopify Stores in the UK for Inspiration