Is Your Store Ready for AI Shoppers?

Something changed in how people shop this year, and most store owners have not felt it yet because it does not show up as a new traffic source in analytics. A growing share of buyers now start with an AI assistant — ChatGPT, Gemini, Perplexity, or the assistant built into their phone — and ask it a question instead of typing into a search box: "find me a waterproof jacket under £150 that ships to Manchester by Friday." The assistant compares products across stores, shortlists two or three, and increasingly can complete the purchase inside the conversation. If your store is not one of the options it considers, you were never in the running, and you will never know.

This is not the same as SEO, though it rewards some of the same discipline. A search engine ranks pages; an AI shopper reads products. It wants to know exactly what the item is, what it costs today, whether it is in stock, what it will cost to deliver and when it will arrive, what the return terms are, and whether other buyers were happy. Then it decides. Stores that give it those answers clearly, in machine-readable form, get picked. Stores that hide the price behind a click, load the stock status with JavaScript after the page renders, or describe a product as "premium quality" without saying what it is made of, get skipped. Here is the checklist we now run on every e-commerce build and audit.

One: complete, structured product data. Every product page should carry Product schema with name, brand, price, currency, availability, condition, ratings and identifiers such as GTIN or MPN where they exist. This is the single most important item on the list, because it is the data AI assistants read first. Most platforms emit partial markup by default; check yours with a validator and fill the gaps.

Two: prices and stock in the page itself. If the price or availability only appears after a script runs, an assistant reading the raw HTML sees a product with no price — and a product with no price does not get recommended. Render them server-side, keep them accurate, and update the structured data when they change.

Three: plain, specific product descriptions. Write for a careful buyer who cannot see the item: material, dimensions, weight, what is in the box, what it is compatible with, who it is for and who it is not for. Assistants match questions to specifics; adjectives give them nothing to match. This also happens to be what converts humans.

Four: a shipping and returns page that answers every question. Cost by region, delivery time by service, cut-off times, the return window and who pays for return postage. Put it in text, not an image, and link it from every product page. "Ships to Manchester by Friday" is a question your store either answers or loses on.

Five: genuine reviews, on the page and in the markup. Assistants weigh ratings heavily, and they read the review text to answer questions like "is it true to size". Collect reviews after delivery, show them unedited, and mark them up so they count.

Six: a product feed that is current. Google Merchant Center, and now the shopping programmes from OpenAI and others, ingest feeds directly. An accurate feed — matching your site, updated when prices and stock change — is how your catalogue enters those systems at all. A stale one gets your store demoted or dropped.

Seven: a fast, crawlable site. Assistants fetch pages on demand and give up on slow ones. Core Web Vitals in the green, no content behind interstitials, clean HTML, a current sitemap, and robots rules that deliberately allow the AI crawlers you want (check for GPTBot, PerplexityBot and Google-Extended rather than assuming).

Eight: a checkout an agent can complete. This is the newest frontier and the one to watch rather than panic about: guest checkout without forced account creation, standard address and payment fields, clear error messages, and support for the emerging agent-checkout protocols your platform offers. Stores that make checkout hostile to automation will find the assistant quietly completes the sale somewhere easier.

How do you know where you stand? Ask the assistants yourself. Take your ten best-selling products, describe each the way a customer would, and ask ChatGPT, Gemini and Perplexity what to buy and where. Note who they name. Do it monthly. That list is your real competition now, and watching your store appear on it — or not — is the clearest measure there is.

None of this replaces a good store for human visitors; it rewards one. Every item on the list makes the site clearer, faster and more honest for a person too, which is why the stores that do well with AI shoppers are usually the ones that were already doing well. If you would like us to run this checklist against your store and tell you plainly what is missing, that is a one-week piece of work and it is where we start every e-commerce engagement.