The proof
Nothing is decided
after the fact.
Every number GetMOW reports can be recounted by the client, because the way it is counted is agreed in writing before any work starts and never changed afterwards. What follows is the whole of that arrangement, including the part of the tracked set that is deliberately left alone so the rest has something to be measured against. The commercial terms are the ones on the order form and nowhere else.
Agreed before work starts
Eight things are fixed in writing on day zero: the baseline, the target, the engines, the prompt set, the market, the language, the measurement method and the exclusions. None of them is GetMOW’s to change once signed, and success is never something GetMOW gets to define afterwards.
Recountable by the client
The prompt set never changes on GetMOW’s say-so. It is agreed at the start, and after that it moves only when both sides agree in writing that the category itself has moved: a competitor that no longer belongs in the question, or a new one that plainly does. Leaving a dead name in the set measures a category nobody is buying in any more, so the point is not that the questions are fixed for ever, it is that neither side can change them alone or quietly.
Every change is dated and recorded with the reason for it, so any reading can be traced to the set that produced it. Run counts are published. Every run is logged with every source the engine cited or fetched, and the log is the client’s to keep.
A third-party tracker, in the client’s own name, reads the same set beside GetMOW’s system. At each milestone date either side can call a witnessed rerun. If the two readings differ by more than ten points, the lower one governs.
A random set of your questions is left alone
Part of the tracked set is worked and part of it is deliberately not, and which questions fall on which side is settled at random and fixed in writing with everything else on day zero. The untouched questions get the same measurement, at the same seven runs per engine per day, and none of the work.
The gap between what moved in the worked questions and what moved in the untouched ones is the part GetMOW puts its name on. Whatever moved in both moved for reasons nobody here can claim: an engine update, a rival going quiet, attention spilling over from the brand as a whole. Those are recorded in the ledger and claimed by nobody.
A holdout costs GetMOW the easiest half of the story, which is why most vendors do not run one. A number that cannot be told apart from what would have happened anyway is not a result, and there is no way to tell the two apart without leaving some questions alone.
Paid on the result, in that order
Engagements start at €5,000, which buys a 49 day proof phase: the measurement agreed in writing, the category mapped, the plan on the table, and one qualifying mention live on a relevant third-party site that neither the client nor GetMOW owns, at a URL the client can open. If it is not there by day 49, the €5,000 comes back in full.
Past day 49 the implementation is quoted before it starts, and the two success payments across the first year fall due only once the target attached to each has been verified against the measurement agreed on day zero. A target that is missed is waived, not invoiced.
If a material engine change causes the agreed source set to lose more than half its citation share in a quarter, either side can call one 60 day pause while the map is rerun and both sides agree the new set. The milestone dates move by the length of the pause and nothing else moves with them.
Questions about the proof
- How do you know the change was the work and not the engine?
- Because a random part of the tracked set is deliberately left untouched. GetMOW works one part of the agreed questions and leaves the other alone, measures both the same way at seven runs per engine per day, and claims only the gap between them. An engine update, a rival going quiet or attention spilling over from the brand as a whole moves both halves together, so it is recorded in the ledger and claimed by nobody.
- Who controls the measurement?
- Nobody controls it after the fact, which is the point. The baseline, the target, the engines, the prompt set, the market, the language, the measurement method and the exclusions are fixed in writing before work starts. GetMOW logs every run and every source, a third-party tracker in the client’s name reads beside it, and at each milestone either side can call a witnessed rerun.
- What happens if the two readings disagree?
- The lower reading governs. GetMOW’s system and the client’s own third-party tracker read the same agreed prompt set, and when the two differ by more than ten points at a milestone the lower number is the one the milestone is judged against. A number GetMOW cannot reproduce in front of the client is not a number GetMOW gets paid on.
- What does the guarantee not cover?
- The guarantee covers two dated commitments: the qualifying mention by day 49, refunded in full if missed, and the two success payments, each waived if its target is not verified. It does not cover the implementation from day 50 or the monthly work in the second half of the year, which are billed for the work performed, and it promises nothing about first place, revenue, click-level attribution or questions outside the agreed set.
- Is one free report enough to judge a category?
- No. The free AI visibility report is a single day’s read across three AI surfaces, and the answers move between runs, so it is a measurement with a date on it rather than a permanent verdict. It is enough to show whether a company is absent and who is named instead, which is what a buyer needs in order to decide whether the engagement is worth a proof phase.
Sources
This link is to the terms for this website and the free AI visibility report, read in September 2026. The figures above are the order form’s: a client reads it before signing, and it is not published.
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.