The questions
Your buyers write
the questions.
The evidence decides
which ones qualify.
GetMOW tracks the questions a buyer asks before they know your name, not the ones a brand would like to win. The set is written from what your own team hears, tested on every engine before anyone commits to it, and agreed by both sides. A question enters the tracked set only if it passes four tests, and every one of those tests is settled by a run rather than by an opinion.
Where the questions come from
Week one is a call with your team: the people who hear buyers first. SDRs, customer support, account executives, anyone who sits in front of a customer often enough to know how the questions are actually phrased. What comes out of that call is the raw material, the problems buyers describe, the words they use for them, and the objections that come back often enough to have a name.
From that, GetMOW writes 150 to 200 questions in the buyer’s own phrasing. A buyer with a problem does not type your product name, so most of the set never mentions it: they are questions about the problem, the category and the alternatives. A branded set is added on top, the questions a buyer asks once they have heard of you: “[Client], are they any good”, “[Client] reviews”, “[Client] vs [rival]”.
What makes a question qualify
Every candidate is run seven times on every engine in scope, logged out, in your buyers’ language and country, before anyone decides whether to track it. Writing a question is a guess. Running it is evidence, and the evidence is what settles the set.
Four tests, all of which have to pass. The rule is deliberately two-sided: it stops GetMOW picking easy questions to look good on, and it stops anyone picking vanity questions no buyer asks. You see the full run behind every decision before you sign, and both sides sign the list.
- A buyer actually asked it
- The question came out of the call with your team, or out of the way buyers phrase it in the wild. A question nobody asks is a question nobody reads the answer to.
- You are not already in the top three
- A question you already win is a screenshot rather than work. Tracking it would show movement that was there before anyone was paid for it.
- At least one rival is being recommended
- The answer has to name somebody. A question the engines answer without recommending any vendor has no shortlist on it to get onto.
- The category is not settled by a single giant
- Where one or two names hold the answer across every run, six months of work does not move it. Those questions are screened out before you sign rather than explained away afterwards.
What being in the answer means
One run is one prompt, on one engine, on one day. Present means the answer names you as an option for the buyer: not in passing, not as an example of something else, and not as a warning. A mention is not a recommendation, and the two are counted separately.
Position is your place among the distinct vendors an answer recommends, in the order they appear. Distinct is the word that does the work: an answer naming the same vendor three times has recommended one vendor, not three.
No figure is published off a single run. A rank that appears on Tuesday and is gone by Friday belongs to the engine rather than to anything done that week, and GetMOW does not show a number with fewer than seven runs behind it.
Questions about the questions
- How many prompts does GetMOW track?
- Between 150 and 200 questions are written for a typical engagement, in the buyer’s own phrasing, plus a branded set of the questions a buyer asks once they have heard of the company. Not all of them are tracked: every candidate is run seven times on every engine first, and only the ones that pass the qualification tests enter the set that is measured and worked.
- Who decides which prompts are tracked?
- Both sides, on evidence neither side can adjust afterwards. The questions come from a call with the client’s own customer-facing team, every candidate is run seven times on every engine before it qualifies, and the client sees the full run behind each decision before signing. Both sides sign the resulting list, and after that it moves only when both sides agree in writing that the category itself has moved.
- Can GetMOW pick easy prompts to make the results look good?
- No, because a question the client already wins is disqualified. A prompt enters the tracked set only if a buyer actually asks it, the client is not already in the top three, at least one rival is being recommended in the answers, and the category is not settled by a single giant. The same rule that blocks easy questions also blocks vanity questions no buyer asks.
- What does it mean to be present in an AI answer?
- Present means the answer names the company as an option for the buyer, rather than in passing or as a warning. Position is the company’s place among the distinct vendors that answer recommends, in the order they appear, so an answer naming one vendor three times has recommended one vendor. One run is one prompt on one engine on one day, and no figure is reported off fewer than seven of them.
Sources
The rules on this page are GetMOW’s own. They are the ones written into the measurement agreement a client reads and signs before any work starts, and that agreement is not published. Read in September 2026.
Start with what the answers say about you.
Before you weigh any method, including this one, see the answer ChatGPT, Gemini and Perplexity give today when a buyer asks who to consider in your category. The AI visibility report is free, written for your category, and yours whatever you do next.