Measuring whether AI assistants recommend you
Getting cited by the AI that answers for you
When someone asks ChatGPT, Perplexity, or Google's AI for the best tool in your category, does it name you? This measures that — and shows exactly where you're invisible and how to fix it.
Buyers increasingly ask an AI instead of searching — and most companies have no way to know whether the AI recommends them or never mentions them at all.
A system that measures whether the major AI assistants name you when they answer questions in your category, scores where you stand against competitors, and ranks the gaps to fix first.
| Before | After | |
|---|---|---|
| What gets measured | where you rank on a Google results page | whether the AI names you when it answers |
| How you'd check it | ask the AI a few times by hand | a scored, repeatable read across many questions |
| Knowing where you stand | a guess | a health score and share against named competitors |
| Knowing what to do | no clear next step | gaps ranked by effort against payoff |
| Fixing the gaps | write pages and hope | drafts built the way AI assistants prefer to read |
This is a whole measurement discipline that didn't exist two years ago, stood up as a repeatable system: a clear baseline of where you stand now that AI answers the questions, and a prioritized list of gaps ranked by how much work each takes against how much it's worth. The job isn't 'rank higher on Google' — it's 'show up at all when the machine answers,' which most companies can't yet even measure. Being the answer the AI gives is the kind of position that compounds, and the companies that start measuring it now build a lead while everyone else argues about whether it counts.
What I built
A pipeline that treats AI answers as a new place you can be found — and measures and improves your standing there, the way search marketing once measured Google rankings.
- Grounded in real questions. It checks whether the AI names the product across a careful list of high-value questions buyers actually ask in the category — so the picture rests on real queries, not a hunch.
- A clear read of where you stand. It scores the product on a small set of health metrics and measures how often the AI names you versus named competitors — a plain answer to "who do the machines recommend, and how often."
- Effort aimed where it pays. It ranks the gaps by buying intent, so work goes to the answers most likely to turn into a customer rather than the ones that just look bad.
- Drafts the fix. It writes the pages to close those gaps, structured the way AI assistants prefer to read them — with blanks marked wherever a real fact belongs, so the model organizes the page without making the specifics up.
The whole thing is built to be run again and again, so the picture stays current as the AI assistants change their answers.
Why it matters
The payoff is being the answer, not just a search result. When a buyer asks an AI for the best tool in your category and it names you, that's the position that wins the deal — and until now there was no way to even tell whether you held it. This makes it measurable: a company can see itself the way the AI sees it, find exactly where it's invisible, and fix it deliberately.
Normally, a brand-new measurement discipline takes years and a whole industry to professionalize. Here it's a working system standing up in real time — and the companies that build the baseline now compound a lead while everyone else is still arguing about whether AI answers "count."
The hard part is that there's no playbook for this yet — it's a discipline being invented in real time. Asking an AI a question once tells you nothing reliable; the answers vary, and a few hand-run checks can't be trusted. So the real work is choosing which questions to track (the ones real buyers actually ask), deciding what "showing up well" even means as a set of metrics, and ranking the gaps by how close each question is to a purchase. The other discipline is in the drafted fixes: pages are written with blanks marked where each real fact belongs, so the AI organizes the page without inventing the specifics. That restraint is what makes the output safe to publish.
The bottom line is that a company can now see itself the way the AI sees it — and act on it deliberately instead of guessing. Where old search marketing took fifteen years to professionalize, this is a measurable discipline being built in real time, and the lead goes to whoever starts measuring first.
Which questions to track, the handful of metrics behind the health score, and how the gaps get ranked by buying intent are what separate a real gap from noise — and the method clients pay for.