You do not need a consultant to find out whether AI search engines know your business exists. You need thirty minutes, an internet connection, and a willingness to read what comes back.

The five experiments below are the ones we run on every prospect before a discovery call. They take a few minutes each. They cost nothing. And they will tell you, with surprising clarity, whether your business is currently visible, partially visible, or invisible in AI search.

Think of this as the Mythbusters episode for your own marketing. Run the tests. Believe the results.

Experiment 1: The "name me" test

Setup: Open ChatGPT. Type the exact query a customer would type. For example: "Who is the best [your service] in [your city]?"

What to look for: Does ChatGPT name your business? If yes, where in the list? If no, who does it name?

What to do with the result:

  • If your business is named first or second, you are winning. Do not get complacent.
  • If your business is named third or later, you are visible but beatable. Tighten schema and GBP.
  • If your business is not named, you are invisible. Move to experiment 2.

Bonus round: run the same query in Perplexity, Google AI Overviews, and Gemini. Compare. The four answers will not be identical, and the differences are diagnostic.

Experiment 2: The "tell me about" test

Setup: Open ChatGPT. Type "Tell me about [your exact business name]." Use the name as it appears on your website.

What to look for: Does ChatGPT know who you are? Does it correctly describe your services? Your location? Does it have your website URL right?

What to do with the result:

  • If ChatGPT confidently and correctly describes you, your entity signals are strong.
  • If ChatGPT says "I don't have specific information about that business," your schema and citations are too weak.
  • If ChatGPT describes you incorrectly (wrong city, wrong services, wrong owner), you have an entity confusion problem. This is fixable but it requires consistent NAP work across every directory.
The "tell me about" test is the single most important experiment. If AI cannot describe your business accurately, it cannot recommend it confidently.

Experiment 3: The schema sniff test

Setup: Go to the Google Rich Results Test (search.google.com/test/rich-results). Paste your home page URL. Hit "Test URL."

What to look for: A green "Page is eligible for rich results" message and a list of detected schema types (LocalBusiness, Service, FAQPage, etc).

What to do with the result:

  • If you see LocalBusiness, Service, and FAQPage detected, you are ahead of most of the home services market.
  • If you see "no rich results detected," your site is invisible to AI engines that rely on structured data. This is the highest-leverage fix you can make.
  • Repeat the test on three of your service pages. Schema only on the home page is not enough.

Experiment 4: The llms.txt check

Setup: In your browser, type yourdomain.com/llms.txt and hit enter.

What to look for: A plain text file that lists your core pages and describes your business in a model-friendly format.

What to do with the result:

  • If the page loads with content, you are in the early-mover club. Skim it for accuracy.
  • If you get a 404 error, your site has no llms.txt. This is a one-day fix that compounds for years. See our explainer on what llms.txt is and why your site should have one.

Experiment 5: The NAP audit

Setup: Open a fresh browser tab. Search Google for your business name. Click the GBP listing. Note the address and phone number exactly. Now open your website. Note the address and phone number exactly. Now check Houzz, Yelp, Bing Places, and any industry directory you remember signing up for. Note each address and phone.

What to look for: Six to ten records of your business across the web. Are the name, address, and phone number identical, character for character, on every one?

What to do with the result:

  • If everything matches, you are stronger than most. Move on.
  • If you find variations ("123 Main St" vs "123 Main Street," "(555) 555-5555" vs "555.555.5555"), you have NAP inconsistency. AI engines treat this as low confidence. Fix every record to match one canonical version.
  • If you find directories you did not know existed, claim them or update them. Old directory listings with stale info are worse than no listing.

The thirty-minute scoreboard

After five experiments, you have a clear-eyed picture of your AI visibility. Use this scoreboard:

Score yourself
  • 5/5 passes: You are doing the work. Keep the cadence steady.
  • 3 to 4 passes: Solid foundation, real gaps. Pick the failing tests and fix them in order.
  • 1 to 2 passes: Most home and design businesses live here. The wins are huge if you act on them.
  • 0 passes: Invisible. The good news is the cheapest, fastest improvements are still ahead of you.

What to do next

If you ran the experiments and want a second opinion on the results, that is exactly what our free Quick Audit is for. We run the same five tests, plus a few more we do not publish, and email you a one-page report with the order to fix things in.

If you want the broader playbook, read the 8 AI visibility signals every home and design website needs in 2026 and our 2026 GBP optimization checklist.

The goal is not to be perfect. It is to know where you stand. Run the experiments. Believe the data. Then fix what is broken.