// Pricing
What it costs
One fixed price audit and four retainers, published in full. You can work out whether this is affordable before you speak to anyone.
Start here. One price, one deliverable.
The Audit is a complete piece of work with an end date. It is not a sales document and it does not need a retainer behind it to be worth having.
// One off
The AI Visibility Audit
$7,500fixed
Three weeks from kickoff. No call required. The scope does not move and neither does the price.
- Demand map of the questions that drive your revenue
- Every gap, engine by engine, across ChatGPT, Perplexity, Claude, Gemini and AI Overviews
- Every error about your brand, traced to the page that taught it
- Nine dimension readiness score with the blockers named
- 90 day roadmap ranked by impact, yours to keep
- Verified fact sheet for your brand
Credited against month one. Sign a retainer within 60 days and the audit fee comes off your first invoice, up to the value of that first month.
Get the Audit →Four retainers. One engine.
Every tier runs the same engine and moves along the same five axes: markets, tracked questions, how often the set reruns, briefs, and third party targets. Nothing else changes between them, which is why you can read the difference off the card.
-
The measurement half
GEO only
$6,500/mo
For a team that wants the AI answer engine work and already runs its own search program.
- One market
- 60 tracked questions, rerun every quarter
- No content briefs
- 8 outreach targets a month
- 90 day minimum, month to month after
-
Both halves, one market
Foundation
$7,500/mo
The entry point to the full engine. Search and AI answers run as one program in your primary market.
- One market
- 60 tracked questions, rerun every quarter
- 4 content briefs a month
- 8 outreach targets a month
- 90 day minimum, month to month after
-
Four markets, read monthly
Compound
$15,000/mo
For a company selling in several markets that needs movement read every month rather than every quarter.
- Up to four markets
- 120 tracked questions, rerun every month
- 8 content briefs a month
- 20 outreach targets a month
- Six month minimum
-
Every market you sell in
Category
$30,000/mo
For the brand that intends to be the answer in its category, in every market it sells in.
- Every market you sell in
- 240 tracked questions, rerun every month
- 16 content briefs a month
- 40 outreach targets a month
- Twelve month minimum
Every tier includes the same measurement. A fixed question set built with you and then held constant, put to five answer surfaces, ChatGPT, Perplexity, Claude, Gemini and AI Overviews, at least five times each because one run of a non deterministic engine is not a measurement. Every finding is read by a person before it reaches you. What we refuse to claim from those numbers is set out on how we measure.
Foundation and GEO only rerun the full question set every quarter. Compound and Category rerun it every month, which is the reason the question counts and the prices move together.
SEO alone is quoted at the Foundation price. The search half cannot be run cheaply against a baseline that has not been measured, and measuring it is the part that costs.
Extending a tier.
The four things that can grow past your tier, and nothing else. Each one is quoted against your tier and agreed in writing before the work starts, because the real cost is machine time and review hours and both are computable for your set rather than guessable from a list price.
- Extra market
- Each market brings its own question set, its own briefs and its own outreach list. Nothing about a market is shared with another one, which is why it is priced rather than absorbed.
- More questions
- The tracked set can grow past the tier. We compute the machine time and the review hours for your specific set and quote from that, because those are the two things that actually move.
- More briefs
- Additional content briefs beyond the tier, agreed in blocks at the start of a quarter so your writers are never waiting on us.
- More outreach
- Additional third party targets a month, worked exactly the way the ones inside your tier are worked.
The three ways the price moves.
All three are published. Any combination of them is written into your agreement before you sign, never assumed.
30%
Founding clients
Thirty per cent off the first six months. Three slots. In exchange, a named case study once the numbers are in. They close when they are gone.
10%
Annual prepay
Ten per cent off when the year is paid up front. It buys us a year of certainty and it buys you a year at one number.
$7,500
Audit credit
The audit fee is credited against month one of any retainer signed within 60 days, up to the value of that first month.
What is never included.
Not a list of upsells. These are the things we do not do at any price, so you can put them in someone else’s budget line before you sign ours.
Paid media
We do not run ads or take a percentage of your spend. Nothing we recommend should depend on us also being paid to buy the traffic.
Development work
We specify the fix and verify it live. Your engineers ship it, or your agency does. We do not bill for sitting in their sprint.
Community and KOL
Not our discipline. Anyone who tells you a paid thread is a citation strategy is selling you a number that will not survive a rerun.
Token launch marketing
A launch campaign and an answer engine program are different jobs on different clocks. We do the second one.
Writing your content
We deliver strategy, architecture, templates and detailed briefs; your team or your writers execute them. In crypto and fintech that keeps your voice and your compliance review inside your business, where it legally belongs.
Who the invoice pays for.
Ben Tabas
Every engagement is run by Ben Tabas, who owned the organic search channel at a top 5 global crypto exchange. There is no account manager between you and the person doing the work, and no junior running your scans on the quiet.
At these prices the machine time is the cheap part. What you are buying is the hours spent reading what the engines said, deciding which findings are real, and tracing the ones that are back to the page that caused them.
Straight answers.
Why is buying the GEO half alone barely cheaper?
Because the measurement is most of the cost. Building the fixed question set, running it through five engines, and having a person review every finding costs almost the same whether or not we also run the search half against it. GEO only is priced at what it costs to deliver, and Foundation adds the search program on top of a baseline that is already paid for. That is the whole reason the two numbers sit close together.
Do you discount for buying both SEO and GEO together?
No. We do not discount for buying both, and we do not price the halves so that buying them separately is cheaper. A discount for taking both would be an admission that they are two products bolted together, and they are not: they run off one question set, one fact sheet and one roadmap.
What does the audit credit actually do?
The Audit is $7,500. Sign a retainer within 60 days and that fee is credited against your first month, up to the value of that first month. It is a credit against an invoice rather than a payment back to you, so the most it can do is take month one to zero.
Can we pause a retainer?
No. The first term is committed, and after it you are month to month at Foundation and GEO only. The work is cumulative: pages, sources and citations earned in one month are what the next month builds on, and a paused month does not stop the decay, it just stops the building.
What is the minimum commitment?
Ninety days at Foundation and GEO only, six months at Compound, twelve at Category. The minimums rise with the tier because the bigger programs spend their first months on work that only pays back later: more markets, a wider question set, and third party sources that move at the pace of other people’s editorial calendars.
What am I actually paying for at these prices?
Mostly senior time and human review. The engine calls behind a month of tracking cost a few hundred dollars. Every finding is then read by a person before it reaches you, because the automated pass produces false positives and in our own testing at least one finding in eleven did not survive that reading. Software gets you a triage queue. A person gets you a finding list, and that review is most of the invoice.
Why publish prices at all?
Because a category where every competitor hides behind "book a call" is a category where buyers assume the number is bad. You should be able to work out whether this is affordable before you speak to anyone, and if it is not, we would both rather know now.
Know the number before the call.
Start with the free scan, or start with the Audit and have the whole picture in three weeks.