NamedLocal › Guides › Check your AI visibility
When a customer asks ChatGPT or Google's AI for the best plumber, dentist, or accountant in town, the answer names a handful of businesses and skips everyone else. Here's how to find out which side you're on, using nothing but the free engines and a notepad. It's the same method we use in our published studies.
Step 1
Ten is enough. Use the customer's words, not yours: nobody types "premier full-service plumbing solutions." These are the exact question shapes we use in our studies; swap in your trade and city:
Include the city and state every time. "Near me" without a city tests the engine's guess about the asker's location, not your visibility.
Step 2
Three engines cover most real usage: ChatGPT, Google's AI (AI Mode, or the AI Overview at the top of a normal search), and Perplexity. Two rules that keep the result honest:
Step 3
This is the step everyone skips, and it's the one that separates a real check from an anecdote. AI answers move: re-run the same question minutes later and the named list shifts. In our studies we ask every question multiple times and only count a business as visible for that question if it's named in the majority of runs. Do the same, even if your second pass is a different day. A single run can flatter you or damn you by luck.
Step 4
The competitor list is the most useful thing you'll write down. Those businesses have a footprint the AI can read, and yours is the same market, so the gap is findable. For each name that keeps coming up, check: which directories list them, which "best [trade] in [city]" articles name them, and whether their homepage says what they do and where in plain text. National chains showing up (Roto-Rooter appeared in every city of our plumbers study, ClearChoice in four of six dental markets) tells you the answers reward heavy web footprints, not local reputation.
Step 5
| Named in most answers, most runs | visible | Rare. Defend it: keep listings consistent and your site's plain text current. |
| Named sometimes, or only for some questions | partial | Fragile. A phrasing change or a competitor's improvement can push you out. |
| Named in none | invisible | Roughly half of established local businesses in our studies. Fixable, but not by waiting. |
If you land in the bottom row, the usual causes are boring and concrete: a Google Business Profile that's thin or miscategorized, absence from the "best of [city]" roundups the engines quote, inconsistent name/address/service info across directories, and a homepage that never states the service and city in text a machine can lift.
Questions owners ask us
No, and this is the most common wrong assumption. In our national plumbers study, Denver's most-reviewed plumber (about 3,300 reviews) was named in zero of 20 answers, while shops with a tenth of the reviews got named. Reviews are a signal for humans; AI assembles answers from a different footprint.
Monthly is enough for most businesses. Answers drift on their own, and if you've made changes (site copy, listings, new mentions), give them several weeks to show up. Perplexity tends to reflect changes fastest and ChatGPT slowest.
Schema helps a machine read what's already on the page. It won't rescue a site that never says plainly what you do and which cities you serve. Fix the plain text first, then add the markup.
No. Our studies keep finding businesses that rank at the top of Google Maps, with awards and thousands of reviews, that AI never names. The two are assembled from different sources, so you have to check them separately.
The data behind this guide
53% of 74 established plumbers across 8 US cities never named by AI.
49% of 37 award-listed dental practices across 6 US cities never named by AI.
Same method, more markets.
Questions about the method, or want your market run as a study? Email jeff@namedlocal.com.