Can an AI assistant read, trust and recommend your store?
A growing share of product discovery now starts inside ChatGPT, Gemini and Perplexity rather than in a search results page. Most Shopify stores are close to invisible to all of them, not because the products are wrong but because the structure, the data and the content were built for something that reads differently.
Put your URL in. You get a scored report in about two minutes. No call, and no follow-up sequence unless you ask for one.
What it checks.
Whether you can be read at all. Product, price, stock and returns information that exists only inside images or scripts is invisible to the systems doing the recommending.
Structured data. Product, offer, review, question and organisation markup. Present, valid, and actually consistent with what the page says, which is where most stores fall down.
Whether you answer the question. Content that addresses what people ask before buying, in a form a model can quote back rather than paraphrase badly.
Trust signals. Reviews, returns policy, delivery terms and a real contact route. A recommendation engine checks these before it puts its name to you, much as a customer would.
Crawl and access. Whether you are blocking the crawlers that feed these systems, on purpose or by accident. Accidentally is more common.
Ordinary search health. The fundamentals still matter and the report scores them alongside the newer things, rather than pretending Google has stopped counting.
What comes back.
A score out of a hundred, broken down by the six areas above, with every failure explained in plain English and ranked by what it is likely to be costing you. Where a fix is a ten minute job it says so. Where it is a project, it says that too, and roughly what kind of project.
There is no obligation to speak to me afterwards. If the report is useful on its own, take it and go. Plenty of people will, and that is fine.