You rebuilt your homepage. New hero, cleaner sections, a product grid you are actually proud of. It looks good on every device, and the numbers held steady.
Then a shopper asks ChatGPT which store sells the thing they want, and your store is not in the answer.
Not because your page is slow. Not because your copy is weak. Because the thing that read your page never saw most of it.
This is a new failure mode, and it is quiet in the way the worst problems always are: nothing breaks, nothing alerts, the traffic just goes somewhere else.
What an AI assistant actually reads
When a shopper asks an assistant for a recommendation, the assistant is not browsing your site the way a person does. It is working from pages a crawler fetched earlier, or fetching one live, and reading whatever came back.
Here is the part that surprises most merchants: as of mid-2026, the major AI crawlers do not run JavaScript. GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot — they request your page, read the HTML the server sent, and stop.
Google is different. Google's crawler does execute JavaScript, which is why a store can rank fine in normal search and still be missing from AI answers. Merchants read that as a contradiction. It is not. The two systems are reading two different versions of the same page.
So the question is not "does my page look right in a browser". The question is "what is in the HTML before anything runs".
Why your Shopify homepage fails this test first
Of all your pages, the homepage is the one most likely to be assembled by scripts.
Think about what a modern Shopify homepage actually contains:
- a slideshow that loads its slides after the page does
- a featured-collection block that fetches products on scroll
- a reviews widget from a third-party app
- a currency switcher that rewrites every price
- a cookie banner sitting on top of all of it
Most of that arrives after the initial HTML. A person sees a full page because their browser runs all of it. A crawler that does not run JavaScript sees the skeleton — a nav, a footer, a few headings, and a lot of empty containers where your products were supposed to be.
The gap is easy to underestimate because you never encounter it. You look at your store in a browser, the browser runs everything, and the page is complete. There is no view in your admin that shows you the stripped-down version.
Your product pages are usually in better shape, because Shopify themes render most product content server-side by default. The homepage is where merchants stack the most apps, and it is also the page that carries your positioning: who you are, what you sell, why you and not the other store. That is exactly the material an assistant needs to decide whether to mention you.
Four things that make a page unreadable to a machine
Facts that live only inside images. Specifications, sizing, shipping terms, the guarantee — if these are baked into a graphic, they are invisible to something reading text. Your designer put them in the image because it looks better. A parser sees a file name.
Structured data injected after load. Structured data is the block of machine-readable facts that tells a system your page is a product, its price, whether it is in stock. Analysis of AI-cited pages found 65% of those cited by Google's AI Mode and 71% of those cited by ChatGPT carry it. But if an app writes that block with JavaScript after the page loads, the crawler that does not run JavaScript never receives it. You have the markup. It just is not there when it counts.
Crawlers blocked in robots.txt. Some stores block AI crawlers, sometimes deliberately, sometimes because a plugin added a blanket rule. If OAI-SearchBot cannot fetch the page, nothing else on this list matters.
Copy written for scanning, not for answering. "Premium quality. Designed for you. Built to last." A human skims that and moves on. A machine trying to answer "is this good for cold weather" finds nothing to quote. Pages that get cited tend to answer specific questions in plain sentences.
How to check your own store in ten minutes — or one
There is a manual way, and it is worth understanding even if you never do it.
You open your homepage and view the raw page source — not the inspector panel, which shows you the page after scripts have run, but the actual HTML the server sent. Then you read it. You hunt for whether your product names are in there at all. You look for a structured-data block and check it is not being written in later by an app. You search for your guarantee and your shipping promise as text rather than as pixels inside a banner. Then you open robots.txt and look for rules blocking the crawlers you want.
That is one page. Now do the collection pages. Then every product page, each of which can fail in its own way.
It is an afternoon of reading markup, and at the end you hold a pile of yes-or-no answers with no idea which one is costing you the most money.
This is exactly the kind of work worth handing to a machine. Omnise AI Page Audit reads your whole store the way a crawler would, automatically, in about a minute. It scores three numbers — Performance, SEO, and AEO — and that third one is precisely what this article is about: whether machines can read and quote your pages. Then it ranks every problem it finds by what that problem is costing you, so what you get back is a queue of work in priority order, not a pile of warnings.
The scan is free, it runs on the store you already have, and it finishes while you make coffee.
If the work lands in the "rebuild the structure" pile and you do not have engineering time, the Omnise Expert Crew does it on your store instead of handing you a report. On paid plans the team optimises Performance score to the maximum achievable on your store, and guarantees an SEO score of 100 and an AEO score of 85+. Miss any of it and you get 100% of your plan fee back.
What changes when a page becomes quotable
Fixing this is not a redesign. In most stores it is three unglamorous jobs: move the facts out of images and into text, make sure structured data ships in the server HTML rather than being written by a script, and stop blocking the crawlers you want.
None of that changes how your store looks. That is the strange part — the page a human sees can stay exactly as it is.
What changes is that your store becomes eligible to be mentioned. An assistant cannot recommend a store whose products it could not read, and it will not tell the shopper it skipped you. There is no bounce to look at, no session that ended badly. You simply were not a candidate.
That is the same shape as every other quiet leak: a store that looks fine and still loses money on every visit. The difference is that this one is new enough that most of your competitors have not looked either.
Which is the whole opportunity. The stores that fix this now are not competing against everyone — they are competing against a field that has not checked yet.
You do not have to spend an afternoon in your own markup to find out where you stand. Run the scan, read the list in the order it gives you, and start at the top.
