Two Quick Visibility Tests

Open a new tab. Go to ChatGPT. Ask it about what your business does.

What comes back?

That question takes sixty seconds to answer, and it tells you more than any strategy talk could. Everything about GEO starts from what you just saw, or did not see. A second, related search takes about thirty seconds more and explains why you saw it.

Test One: Ask ChatGPT Directly

Open a new tab, go to chatgpt.com, and skip the account setup. Basic searches do not need one.

Now type a question the way an actual customer would ask it. Not your business name, the problem you solve.

If Neal ran his custom furniture workshop in Denver, he would not search "Neal's Furniture." He would search something like "who builds custom walnut dining tables in Denver," because that is the kind of question a real customer types.

Search for your version of that question now, then read the answer slowly.

• Did your brand show up anywhere in the response?
• Were you cited as one of the sources?
• Did the AI mention you by name at all?

Take a screenshot of what you see, because you are going to want to compare it later. On Windows, press Windows + Shift + S. On Mac, press Command + Shift + 4.

Save it somewhere you will actually find again, like a dated folder in Google Drive, not your desktop where it will get buried by next week. If ChatGPT is slow or at capacity, Gemini works the same way.

ExpandThree steps: ask a real customer question, read whether your brand appears, then screenshot and save the result as your baseline.

There are only two outcomes here, and both of them are useful.

Your brand showed up. That means you already have some AI visibility, even if it is thin. The next question is not whether you exist to AI search but how to strengthen a signal that already works.

Your brand did not show up. That is the outcome for most businesses that try this test, and it is not a verdict on your business. It is a starting point.

An AI that cannot find your name is not making a judgment about your quality. It simply has not found enough evidence yet to build an answer around you. The second test shows you exactly what that evidence looks like.

Test Two: Check Your Web Footprint

Would a stranger trust your business based only on what your own homepage says about it? Probably not, and neither does AI. It checks what everyone else is saying about you first.

Press mentions, guest posts, directory listings, podcast appearances, forum threads: anywhere your brand comes up that you do not control. That collection is what an AI engine reads when it is deciding whether you are worth citing.

Here is how to see your own version of it. Go to Google and search your brand name in quotes, followed by an operator that excludes your own site.

For Neal's furniture workshop, that would look like:

The brand name sits inside quotes. Then a space. Then a minus sign touching site: touching the domain, with no stray spaces breaking up that last part.

If the search returns nothing useful, check for an extra space first. That is almost always the culprit.

ExpandPress mentions, directory listings, forum discussions, and podcast features all feed into the reputation file an AI reads, separate from your own website.

This search shows every place your brand appears on the web except your own site, and you will land in one of two situations.

Thin results. A couple of directory listings, maybe one press mention from two years ago, maybe nothing at all. This is exactly why Test One came back blank. AI could not find enough outside validation to trust you as a source, because there was not enough of it to find.

Rich results. Articles that quote you, a podcast episode featuring your business, forum threads where people mention your name unprompted. If this is what you see, you already have a foundation, and the work ahead is strengthening what exists rather than starting from zero.

Why These Two Tests Are the Same Story

Test One shows you the symptom. Test Two shows you the cause.

If ChatGPT had nothing to say about your industry, this second search usually shows you why: there was not enough third-party evidence out there for the AI to feel confident. That is not a dead end. It is a map of exactly what needs to be built.

Compare this to Neal's situation. If a Google search for his workshop turns up nothing but his own site, ChatGPT has no outside confirmation that his business is real, active, or good at what it does. If that same search turns up a local news feature and two customer testimonials on a woodworking forum, the AI already has something to build a citation from.

Every chapter after this one is about changing what these two searches show the next time you run them. Understanding the stages between invisible and recommended is where that path starts to take shape, but none of it means anything without the screenshots you just took as a reference point.

The Essentials

1. Test One: search a real customer question on ChatGPT or Gemini, not your business name, and see whether you show up.
2. Test Two: search "your brand name" -site:yourdomain.com to see every mention of you on the web that you do not control.
3. The two results explain each other. A blank Test One usually means a thin Test Two. Both are information, not a verdict, and both are worth a dated screenshot to compare against later.

Further Reading and Watching

• Why Your Brand Is Invisible in AI Search and What Marketers Must Fix by CX Today: a look at why AI assistants skip over otherwise-solid businesses.
• What are AI brand mentions? And how are they different from citations?: Yoast's explanation of how unlinked mentions still teach AI models to associate your brand with a topic.