📊 What's Actually Happening

Two checks sound alike and test different things.

The first is the easy one to run: ask an AI engine about your company by name. That tests whether an engine can describe a brand it has been handed.

The second starts without your name. A buyer describes a need, a company size, maybe a budget, and asks for options. That tests whether your brand appears in a list that was assembled before anyone typed it.

A good result on the first tells you little about the second, because the first hands the engine the hardest part of the job, which is deciding who to talk about.

A quieter distinction sits inside the second check. Some responses can fail or go missing during collection, and those need to stay separate from valid answers that simply didn't mention your brand. If the two are counted together, a collection problem can look like a finding about the brand. A few missing responses don't automatically undermine a result, so the working question is whether enough of the planned responses succeeded to support the number being reported. In our own measurement work, we keep coverage of the planned question set separate from how often the brand appeared among the successful answers.

🎯 This Week's Strategic Lens

An unbranded check can support a narrow statement: on these engines, for these questions, at this time, the brand did or didn't appear, and this is how it was presented.

It doesn't support a cause. A brand could be absent because an engine had little to work with, because the question was worded in a way that favored other vendors, because answers differ from one run to the next, or for reasons nobody outside the model provider can observe. It also doesn't support a claim about revenue. One observation hasn't shown that appearing leads to buying.

The honest unit is an observation with its conditions attached, repeated on the same set of questions to see whether it holds.

Practical Section

Four Steps for an Unbranded Buyer Check

1. Write the buyer's questions without your company name, the way a buyer who doesn't know you would ask them: the need, the company size, any constraint that matters.

2. Run the same set on more than one engine: ChatGPT, Claude, Gemini, and Perplexity.

3. Check that enough of the planned responses succeeded to support what you're reading. Keep failed or missing responses separate from valid answers that didn't mention your brand. Then record whether your brand appeared and how: listed, listed with a caveat, recommended with a reason, or absent.

4. Repeat the same set later to see whether the result holds. Changed wording or new questions are a different test, not a repeat. Write down what the result does and doesn't show.

If you can't say what the result doesn't show, hold off acting on it.

🔔 CLOSING SIGNAL

Asking about your company tests what AI can say. Asking for options tests what a buyer would see. Check the one that matches the question you actually have.