Top three in your category's AI answers within six months of the kickoff workshop, across the prompt set we agree on. Miss it and we work two more months free. Miss it again and we part ways.
This is a term of the engagement, written into the contract. Not a marketing claim, and identical on both plans.
3
Top three
The floor we put our own money behind. Number one is always the target.
6
Months
Counted from the kickoff workshop, measured on the prompt set we agree there.
2
Months free
If month six arrives and we missed. Miss again and we part ways.
Fixed questions. Recurring scans. No moving target.
A guarantee is only as honest as its measurement. Ours is procedural, agreed with you before the clock starts, and every reading is yours to see.
01
The prompt set is agreed with you
In the kickoff workshop we agree the category prompts the work will be judged on. Then the set is fixed. Movement is measured against fixed questions, never a moving target.
02
The same prompts are rescanned on a recurring cadence
Each prompt runs more than once per scan, because AI answers vary between runs. What gets recorded is the pattern across runs, not one screenshot.
03
Six engines, named
ChatGPT, Perplexity, Gemini, Claude, Copilot and Google AI. The engines B2B buyers actually ask.
04
Top three means top three in the recommendation set
Where you place in the set of vendors those prompts return, recorded per engine, per prompt, per scan. You see every reading, not a curated highlight.
Two months free. Then, if we miss again, it ends.
No qualifiers stacked on the number, and no disclaimer under it. A remedy that costs us, a clean exit that protects you.
What it does not cover, and what we never promise
It is not number one
Number one is always the goal. Three is the line we contractually stand behind, because promising the single answer would be promising something no honest agency controls.
It is not a revenue promise
The guarantee covers your position in the answer set, not your pipeline. Nobody can guarantee what buyers do after the engines name you, and we will not pretend otherwise.
It is not click-level attribution
AI answers do not come with attribution data. What you get is the answer set before, the work that shipped, the answer set after, and the sources the engines cite.
It is not free work forever
Miss month six and we work two more months at no cost. Miss again and the engagement ends. A guarantee with no end is not a guarantee, it is a hostage situation for both sides.
Because the outcome is measurable in public.
The scoreboard is not ours
Anyone can ask the engines the agreed prompts and read the answer. The measurement lives in public, which is exactly why we are willing to be judged on it.
The prompt set is agreed, not chosen by us
We do not get to pick easy questions after the fact. The set is fixed in the kickoff workshop, with you in the room.
The remedy costs us, not you
Two months of the same work at no cost is real money out of a two-person company. We would not sign that sentence if we did not expect to hit the number.
We choose who we take on
We take a category only when we believe we can win it, and we say no when we do not. Selectivity is what makes the guarantee possible.
Number one is the target. Three is the floor we put our own money behind. Meaningful movement usually starts around month three.
Hold us to it.
Top three in your category's AI answers within six months, or we work two months free. The call starts with a free audit of your own category, and you keep it either way.