How do I know what AI says about my business?
The only reliable way to know what ChatGPT, Claude, Perplexity or Google's AI Overview say about your business is to ask them the plain questions a real customer would ask, one at a time, and read the full answer, including who else gets named. A single spot check on your own phone is a sample, not a full picture, because assistants can vary an answer by phrasing and by session. The AI-Ready Audit runs this systematically across ChatGPT, Claude and Perplexity, reads Google's AI Overview from the search results in each town, and records every word.
ColabContent LLC publishes this page: a Boston AI consulting firm that builds custom AI systems businesses own outright, starting from a $499 audit that shows what Google and AI assistants say about a business today.
The manual check, and why it is harder than it looks.
The most direct way to see what an AI assistant tells a customer about a business is to ask it the same question a customer would ask. Open ChatGPT, Claude or Perplexity in a new conversation, not one with history about the business already in it, and type the question in plain language: "who's a good electrician in Framingham," "best CPA for a fifteen-person manufacturer in the Merrimack Valley," "who handles slip and fall cases near Worcester." Read the entire answer, not just the first name mentioned, and note every business the assistant names, in the order it names them.
This sounds simple and mostly is, but three things make a single check unreliable as a full picture. First, assistants can give different answers to the same question phrased slightly differently: "best plumber near me" and "who should I call for a plumbing emergency in Hudson" can surface different businesses even though a customer might mean the same thing by both. Second, an assistant's answer can be shaped by a person's account history or location settings, so what one owner sees checking from their own phone may not match what a customer in a different town sees. Third, one question tests one moment in time; assistants update what they retrieve, so a business that gets named today might not next month if something about its listings changes, and vice versa.
None of this means the manual check is worthless. It is a genuinely useful first step, and it costs nothing but a few minutes. It just is not a complete audit, because a complete picture needs the same discipline applied across every assistant a customer might use, across every service the business offers, and across every town it serves, with the exact wording of every answer written down so nothing gets misremembered later.
Key Terms
Prompt: the question typed into an AI assistant. Session: a single conversation with an assistant, which can carry context from earlier in the same conversation. Word-for-word record: capturing an assistant's exact answer text rather than a paraphrase, so nothing is lost between the check and the decision made from it. Money Leak Report: the part of the AI-Ready Audit that turns findings like these into a dollar figure.
Building the real question list.
The most useful questions to test are the ones a real customer would actually type, not the ones a business owner thinks describe the business. Start from the three or four services that make the business the most money, and pair each one with each town it serves. A specialty contractor serving four towns and three service lines has, at minimum, twelve real question combinations worth checking, not one, and each combination can surface a different answer.
It helps to think about this the way a customer actually behaves rather than the way a business owner is used to describing the business. A homeowner rarely types the business's own tagline into an assistant; they type the problem they have, in their own words, sometimes urgently. Building the question list from that angle, rather than from the business's marketing copy, is what makes the resulting check actually useful.
Beyond straightforward "best X near me" phrasing, worth testing separately: comparison questions ("X versus Y, which is better"), urgency questions ("who can come out today"), and price questions ("how much does X typically cost near me"), because assistants sometimes name different businesses depending on which angle the customer is coming from. A business that gets named for the generic "best" question but never for the urgency or price version is missing a specific, fixable piece of content.
Write down every answer word for word, along with the date checked, because the value of the exercise comes from being able to compare the same questions again after making a change. A vague memory of "it mentioned someone else" is not enough to know later whether a fix worked. What the assistants are doing with those answers, and why two of them can disagree on the same question, is covered in how ChatGPT chooses which businesses to recommend.
Why a business eventually needs more than a spot check.
A handful of manual checks is a good way to get a first read on whether there is a problem at all. Turning that first read into a ranked list of fixes worth the owner's time takes more structure: enough questions to cover every service and town combination that actually matters, every assistant a customer might reasonably use, and a way to connect each gap back to what it is likely costing in lost calls or lost jobs, so the owner can decide what to fix first.
That is what the AI-Ready Audit does. It asks ChatGPT, Claude and Perplexity, with Google's AI Overview read from the search results in each town where Google shows one, real customer questions built from the business's own services and towns, and records every answer word for word, including which competitors get named instead. Those findings feed directly into the Money Leak Report, which turns the pattern of gaps into a dollar figure and a ranked list of what to fix first, with step-by-step instructions included either way.
What a real check turns up.
A countertop fabricator in Massachusetts, audited in September 2026 (not named here), would have passed the test most owners run. Asked "is the company a good countertop company," ChatGPT and Claude both described it accurately, five stars and all. Those were the only two questions out of eight where either assistant named it, and both contained the company's own name. Asked what a new customer types, "best countertop installers near" its home town or "granite countertops" in the next town over, both assistants listed three or four other fabricators, and the same names kept coming back: across all the recorded answers the two most-named competitors appeared five and four times each.
The pattern is that the question an owner is likely to test, the one with the company's name in it, is the one the assistants can answer from the company's own listings and reviews. The service-and-town questions, the ones a customer asks before they know any names, get answered from whatever pages state a town and a service plainly, and this company's site stated neither: one headline about "the finest countertops," no town pages, and no repair page although the audit's search pull showed local demand for repair.
Once that gap is on paper, the fix is direct: a page per town that names the services offered there, a repair page, and a re-test of the same eight questions after the pages go live to confirm the gap closed rather than assuming it did. That re-test is what the $499 AI-Ready Audit's free quarterly re-check does.
The first week after reading this.
Before booking anything, an owner can build a real, useful picture in a single sitting. Write down the three or four services that bring in the most revenue, and the towns or industries the business actually serves. Pair each service with each town or segment to build a real question list, not a single generic one. Then run every question through ChatGPT, Claude and Perplexity in a fresh conversation each time, and copy the full answer, word for word, into a document alongside the date.
That document becomes the baseline. Whatever gets fixed afterward, whether it is corrected listings, a rewritten page, or a new service-area page, can be checked against this same list of questions to see whether the specific gap actually closed. Without a written baseline, it is easy to believe a fix worked simply because enough time has passed, which is not the same as knowing it worked.
Related pages.
This page is the how-to for checking your own visibility. The two links below explain why the gaps happen and how the underlying decision gets made, and the audit page explains what a full, paid check adds beyond what you can do yourself.
See why ChatGPT skips a business for what causes the gaps you will find.
See how ChatGPT chooses which businesses to recommend for the mechanics behind the check.
See the AI visibility audit for what the $499 audit checks and what it costs.
Start with the $499 audit.
The AI-Ready Audit is $499. The report arrives within 3 business days as a private link and a PDF, with a 5-minute video walkthrough and a 20-minute call. If it has no value you get the $499 back, and every quarter your AI answers, rankings and money leak are re-checked free.
FAQ
The questions below cover the practical details of testing this yourself: how to phrase a check so it matches what a customer types, why the same question can return a different answer on a different day or in a different assistant, how many questions are enough to trust the picture, and when it makes sense to have the whole set run and recorded systematically instead.
How do I check what ChatGPT says about my business myself?
Open ChatGPT, Claude or Perplexity in a fresh conversation and ask the exact question a customer would ask, such as naming your service and your town, then read the full answer including any businesses it names.
Why might I get a different answer than my customer gets?
Assistants can vary answers by session, by account history, and by how a question is phrased. A single check on one device is a sample, not a full picture.
Do I need to check every AI assistant separately?
Yes, because they draw on different sources and often name different businesses for the same question. ChatGPT, Claude, Perplexity and Google's AI Overview each need their own check.
What should I actually ask when I test it?
The plain questions a real customer would type, one per service and one per town you serve, such as "best plumber in Hudson MA" or "who should I hire to redo my payroll system."
What if the answer names a competitor and not me?
That is the finding, not a dead end. It points to a specific, fixable gap, usually in how consistently your business is listed or how plainly your website states what you do and where.
Is there a faster way than checking it myself, question by question?
Yes. The AI-Ready Audit asks the real questions your customers type, built from your services and towns, across ChatGPT, Claude and Perplexity, with Google's AI Overview read from the search results in each town where Google shows one, records every answer word for word, and ranks the fixes by dollar impact, all within 3 business days.
Next step
Try a few checks yourself, then get the full picture with the $499 AI-Ready Audit: real customer questions across ChatGPT, Claude, Perplexity and Google's AI Overview, every answer recorded word for word, ranked by dollar impact.