What ChatGPT Says About Your Business Right Now
Ask ChatGPT, Gemini, or Claude what your business does. Not a competitor's, yours. Do it today, in a fresh chat, worded the way a stranger would word it: "what does [your company] do."
Most business owners have never done this, and most who finally do are unpleasantly surprised. The answer is usually not wrong in a dramatic way. It's wrong in the quieter way that's harder to catch: a service you retired two years ago, a description that misses your actual differentiator, a location or specialty that used to be true, occasionally a confident sentence about something you've never done. None of it is written maliciously. It's just old, or thin, or borrowed from whichever source happened to be easiest to find.
Why this is happening to you specifically
The mechanism is the same one that explains why Copilot is brilliant on some documents and confidently wrong on others, just pointed outward instead of in. When an AI answers a question about your business, it doesn't know you. It searches, retrieves whatever content looks most relevant to the question, and then writes a fluent answer from what it found. The writing is always polished. The retrieval is where the trouble lives, because retrieval finds what's findable, not what's true.
That means your actual answer, the correct and current one, only reaches a customer if it happens to be the thing the AI found and trusted. If your website is hard for these systems to read, if a directory listing from 2021 is easier to parse than your homepage, if a review site or an old press mention is more prominent than your own words, the AI will happily write a fluent, confident paragraph from whichever source won that race. It has no way of knowing it lost.
The part that should concern you
This isn't a future problem you can plan for later. It's already the first impression for a growing share of the people evaluating you, before a phone call, before a form fill, before they ever click through to your actual site. A prospect who asks an AI assistant "what does this company do" is doing exactly what they used to do by reading your homepage, except now the homepage they're reading was written by something else, from sources you may never have checked.
And the failure is invisible to you by default. Nobody tells you when an AI describes you wrong. There's no bounce notification, no support ticket. The only way to find out is to ask, the way I'm asking you to ask right now.
What to do with what you find
Run the test across the three major assistants and write down, plainly, what's wrong: outdated, incomplete, generic, or actually incorrect. That list is the whole starting point. Fixing it is a matter of understanding why a broken link can erase you from an AI's picture of your business just as completely as a wrong answer does, and making sure the true version is the version that's easiest to find and easiest to read, both for a person and for the systems now reading on that person's behalf.
Your website's job quietly changed. It used to need to convince a human. Now it needs to convince a human by way of an AI that reads first, and most sites, including plenty of very good ones, were never built for that second reader.
Ask the question today. What you find will tell you whether this is urgent.
I run this audit for clients directly: what AI actually says about a business, and why. Get in touch if you'd rather know than guess.
Rosemarie Withee has spent thirteen years helping operations teams get real work out of their software, first Microsoft 365, now AI. She’s written six books for Wiley and builds AI products at Portal Integrators.