The short answer
AI shopping assistants recommend products they can read and trust. To get yours recommended, put every purchase-deciding fact in plain text on the product page, add structured product data, answer the questions shoppers actually ask, and keep those facts the same everywhere they appear.
Most stores have room to improve. Adobe found that individual product pages on US retail sites were only 66% machine-readable on average in early 2026, so about a third of their content was invisible to AI models.
Why it matters
Shoppers who arrive from AI are valuable. Adobe found that visits referred by AI services converted at a rate 60% higher than other traffic in July 2026. We covered what that means for stores in AI shoppers convert better. Is your store ready?.
8 fixes for your product pages
1. Put the deciding facts in plain text
Sizes, widths, fit notes, materials and care. If a fact decides a purchase, it should be a sentence on the page, not only a picture or a downloadable size chart.
2. Don’t hide facts behind scripts
Many AI crawlers read a page only as it first loads. A 2024 analysis by Vercel found that none of the major AI crawlers it studied ran JavaScript, apart from Google’s and Apple’s. Content that appears only after scripts run, like some tabs, review widgets and size pickers, may never be seen.
3. Add structured product data
Google’s merchant listing documentation covers properties such as name, image and offers, plus color, material, size, audience, ratings and return policy. Structured data gives machines a clean, labeled copy of the facts on the page.
4. Answer real questions on the page
Add a short FAQ in your shoppers’ own words: “Does this run true to size?”, “Is it good for wide feet?”, “How long does it take to break in?” Put the answer in the first sentence. For an example, see how we answer the most common boot question in How should cowboy boots fit?
5. Say who it’s for and when to use it
AI assistants match products to situations, like a winter wedding or a long day on your feet. Name the occasions and the kind of shopper each product suits.
6. Keep facts the same everywhere
If a product page, a size chart and a help article disagree about widths or sizing, an assistant may repeat the wrong one. Pick one source of truth and update every copy from it.
We do this for our own clients. When we set up Travis for Alvies, we wrote a sizing answer for every boot that matches Alvies’ own product pages, so Travis answers fit questions the way the store does.
7. Make reviews readable
Reviews carry the details shoppers trust most, like how a product fits and feels. Make sure review text is part of the page, not only loaded later by a widget.
8. Measure AI traffic on its own
Track visits from AI services as their own channel, tag the links you control, and watch which products AI shoppers land on and buy.
After the click
Getting recommended is half the job. Shoppers who arrive from an AI conversation expect to keep talking. An on-site AI shopping assistant that knows your catalog can pick up where they left off. That’s what Travis, the voice-first AI shopping concierge from Lone Star Ascent AI, does for Alvies at meettravis.ai, by voice, text or tap.
Choosing one? Start with How to choose an AI shopping assistant for your ecommerce store.
Sources
- Adobe, AI traffic to retail sites and how machine-readable they are, April 16, 2026
- Adobe Analytics AI referral traffic data for July 2026, reported by Digital Commerce 360, August 19, 2026
- Vercel, The rise of the AI crawler, December 17, 2024
- Google Search Central, merchant listing structured data, checked October 31, 2026
FAQ
How do AI shopping assistants decide what to recommend?
They match what a shopper asks for to products they can read and trust. Complete, consistent product facts in plain text, structured data and clear answers to common questions all help.
Does structured data help AI recommend my products?
It helps machines read your product facts reliably. Google’s merchant listing documentation covers properties like name, image, offers, color, material, size, ratings and return policy.
Can AI crawlers read content loaded by JavaScript?
Often not. A 2024 analysis by Vercel found that the major AI crawlers it studied did not run JavaScript, apart from Google’s and Apple’s. Put key facts in the page itself.
What product information matters most to AI shopping assistants?
The facts that decide a purchase: sizes, widths, fit notes, materials, care, who the product is for and when to use it, plus honest answers to common questions.
What should happen when a shopper arrives from an AI assistant?
Let them keep the conversation going. An on-site AI shopping assistant like Travis can ask a follow-up question, recommend products, check the size and add the right item to the cart.
