Services / prices published, bands and all
Two tiers, one ladder, every price on the page.
You should not need a discovery call to learn what this costs. The numbers below are the numbers you would hear on that call.
How does the engagement ladder work?
Free read, then a paid roadmap at $1,500-2,500 with half credited to month one, then the retainer at $3,500-5,000 a month. No other funnel.
Fig. 01 / the ladder, with credits
Rung 1 / free-$750
The read.
Free if we sent it to you; $750 if you came to us, credited against anything that follows. Your AI-answer position, verified by a human.
Rung 2 / one time
The roadmap. $1,500-2,500, one time.
The full move sequence for your prompts, first moves executed. 50% credits to retainer month one. Take it and run it yourself if you want.
Rung 3 / monthly
The retainer. $3,500-5,000 a month.
The whole delivery program plus weekly measurement. 4-month initial term.
Each rung pays for the next, so the cost of deciding early rounds down. Start with GapCheck.
What does the Tier-1 retainer include?
Grounding-source placement, comparison content for buyer prompts, entity and freshness work, and the measurement ladder, at $3,500-5,000 a month.
The dominant lever is third-party: getting credible pages that mention you into the sources AI answers actually cite. Around it: comparison and capsule content for the exact prompts your buyers ask, an entity consistency sweep in month one, freshness plumbing for real-time engines, and analytics onboarding so AI-referred sessions are measurable on your side. Schema and llms.txt setup ships free in week one, labeled as hygiene, because our own evidence says that is what it is.
Terms worth reading: 4-month initial term, a contractual interventions-live-by-week-4 SLA that extends the clock day for day if we miss it, renewal argued from the metric ladder, and nothing in the contract that traps you past the term.
What is the Head of AI-GTM tier?
Fractional Head of AI-GTM for funded teams: strategy cadence plus one shipped system per quarter. Entry $10-12k a month after a $2,500 assessment.
For funded companies that need the function, a senior operator runs it: strategy cadence plus one productized system deployed per quarter, the same engines this practice runs on itself. Entry band is $10-12k a month, ramping to $15-20k as scope grows. The $2,500 AI-GTM assessment is the first step and credits to month one. One seat at a time: this tier is capped at a single concurrent client, on purpose.
Honest status: the adjacent fractional-executive category is well established. This specific hybrid, strategy plus shipped systems in one seat, is our thesis, and it gets tested in public like everything else here.
What are the founding-client terms?
The first three retainer clients pay $2,500-3,000 a month for the initial term, traded against contractual naming rights and a month-5 step-up.
The discount buys something specific: the right to name you, your numbers, and the before-and-after, with quote approval on your side. At the month-5 renewal the price steps up to the standard band and the naming rights survive. That is how the first named case gets built without inventing one.
How do you measure whether it worked?
On a pre-committed metric ladder: interventions live by week 4, brand-mention share in cited sources next, share-of-voice movement 8-13 weeks in.
The ladder is in the contract, with its clock. Leading: the month-one intervention set live and indexed by week 4. Intermediate: your brand-mention share inside the sources AI answers cite, rising in weeks 2-6 after interventions go live. Outcome: share-of-voice movement across your prompt basket, read 8-13 weeks after go-live, never earlier, because reading it earlier reports noise. If the intermediate metric sits flat at week 8 post-live, we re-scope or pause, in writing. Retainers signed before the experiment's first reads are priced on the conservative branch: if the placement thesis misses, the delivery mix re-weights and the placement budget re-prices, in writing.
What do you refuse to sell?
Per-query AI rank, fabricated case studies, astroturfed threads, and billable schema or llms.txt work. Hygiene ships free in week one.
Also on the refuse list: paid placement "into AI answers", which does not exist and is a scam when offered, and Reddit sock puppets. Community content gets drafted by us, posted by you, disclosed as yours. If a lead's winnable prompt surface is too thin to support an honest measurement basket, we decline the lead and say why.
Where does the proof come from?
Every claim traces to a logged measurement you can inspect. No fabricated case studies. We walk you through the work; you decide.
Every competitor page you read today claims a wall of logos. This one shows you a working instrument pointed at our own tool's day-0 gap and a pre-registered experiment that will publish its result either way. Pick whichever proof style you find harder to fake.