Invisible to AI: why your Shopify store is not showing up in ChatGPT or Perplexity, and how to fix it

For years, being found meant climbing the Google results page. That still matters, but a growing share of buyers now skip the results page entirely. They ask ChatGPT, Perplexity, or Google AI to compare options and recommend one, and they act on the answer. This is generative search, and it has quietly changed what visibility means.

The uncomfortable truth is that AI assistants only recommend what they can read and trust. If your product information is thin, locked inside images, or missing structure, the assistant has nothing to quote, so it recommends a competitor whose content it understands. Industry testing through 2026 has been consistent on this: stores that are not structured for language models are effectively invisible in the fastest-growing part of search.

Generative engine optimization, or GEO, is the practice of making your content easy for AI systems to read, verify, and cite. It is not a replacement for SEO. It is the next layer on top, and most of the work overlaps: clear content, clean structure, and schema markup that spells out what you sell.

A jewelry designer came to us frustrated that their pieces never came up when shoppers asked AI assistants for gift recommendations in their category. The site looked beautiful, but almost everything that mattered to a machine lived inside photography: materials, stone types, sizing, and care. There was no structured product data and very little written detail an assistant could actually quote. We rewrote the product content in plain language, added Product and FAQ schema, and built out the questions buyers ask before a considered purchase. The pieces did not change. They simply became something an AI could read and recommend with confidence.

It helps to understand the difference between being included and being recommended, because they are not the same thing. Shopify now syndicates eligible stores into AI shopping experiences automatically, so inclusion is increasingly the default. Being the store an assistant actually names in its answer is not. That comes down to which store gives the model the clearest, most trustworthy information to work from. Two stores selling the same product can sit side by side in an AI index, and the one with richer, better-structured content wins the recommendation almost every time. It is also why this work rewards patience. Unlike a paid placement you can switch on, being trusted by these systems is earned through consistent, accurate, well-marked-up content, and the brands investing in it now are building an advantage that compounds as more buying journeys begin inside an assistant.

Getting visible in AI search is methodical work, not a trick. A sensible order looks like this:

1. Write for the question, not the keyword. AI answers are built from clear, quotable statements. Content that directly answers what buyers ask is what gets cited.
2. Add and validate structured data. Product, Offer, FAQ, and Organization schema give assistants the vocabulary to understand and trust your store. Our SEO and AI search service is built around exactly this.
3. Make your key facts text, not images. Specs, materials, sizing, and policies buried in graphics are invisible to a machine.
4. Keep your entities consistent. The same brand name, product names, and details across your store and profiles help AI connect the dots.
5. Support it with a fast, well-built store. Assistants favour sources that load and render cleanly, which is as much a development question as a content one.

Google publishes clear guidance on the structured data that underpins all of this in its Search Central documentation. The brands that treat AI search as an extension of good SEO, rather than a separate gimmick, are the ones showing up when it matters.