How can I check what AI says about my business in 10 minutes?

Use a temporary or logged-out chat where you can. Assistants that remember you may answer with your own business in mind, which gives you a flattering result your customers won't see. Then make a simple table with four columns: prompt, assistant, were you named, and were the facts correct.

Replace the trade and city with your own. First, the broad one: "Who is the best HVAC company in Austin?" Second, the urgent one: "Who can fix my furnace today in Austin?" Third, the specific service: "Who installs heat pumps in Austin?"

Then three about you by name. Fourth: "What does [your business] do, and where?" Fifth: "What are [your business]'s hours?" Sixth: "Is [your business] any good?" The first three test whether you get recommended. The last three test whether what's known about you is accurate.

Run each prompt three to five times on each assistant. Named in most runs is a real signal; named once is luck. If the broad prompts skip you but the by-name prompts are accurate, assistants know you and don't yet consider you a strong match. If the by-name prompts are wrong, fix the facts first, because a recommendation built on wrong details can do more harm than none.

Pay attention to which competitors appear and what's said about them. It often reveals the evidence you're missing: a stated service area, emergency availability, or more reviews on a particular platform.

It tells you what happened, not why. You'll see that you're missing, but not whether a crawler is blocked, a listing conflicts, or your service pages are too vague. It also can't tell you how many real customers ask these questions. That takes logs, analytics, and asking callers how they found you.

Keep the table and repeat it monthly, so you can see the trend. If you want the causes as well as the symptoms, our free AI-Readiness Check runs the same kind of test and scores the five dimensions behind it. You can see the format first in our sample readout.