There is no reliable public price sheet for this service, and copying a competitor’s retainer is copying a number whose prompt sample, platforms, evidence standard, deliverables, and implementation boundary you cannot see. Cost gives you a floor. The client’s decision value and willingness to pay set the ceiling. Everything here is about computing the floor honestly.
The three cost drivers that actually move the fee
Observation volume. The audit’s variable cost is mostly evidence capture, and it multiplies: a 30-prompt panel on three platforms with three repeated runs is 270 observations. At two to three minutes each to run, save, and classify, that is nine to fourteen hours before anyone reviews a citation or checks a fact. This is why “just add Perplexity” is never a small favor: one added platform on that panel is 90 new observations.
Evidence standard. An audit that saves exact prompts, full answers, classifications, and fact-ledger comparisons (everything the 47-point audit checklist requires) costs two to three times the hours of a screenshot deck. This is the difference between a defensible audit and a demo, and it is the difference clients are actually paying for.
Scope surface. Pages, competitors, languages, locations, and stakeholder count each add fixed chunks of work: more technical checks, more fact approvals, more findings to write, more meetings.
Estimate hours per workstream, from the work itself:
| Workstream | Small diagnostic | Standard audit | Complex audit |
|---|---|---|---|
| Discovery and fact approval | 2 to 3 | 3 to 5 | 6 to 10 |
| Technical eligibility | 3 to 5 | 5 to 8 | 10 to 20 |
| Prompt design and review | 3 to 4 | 5 to 8 | 10 to 16 |
| Observation and classification | 4 to 6 | 10 to 18 | 24 to 50 |
| Source and competitor map | 2 to 4 | 5 to 8 | 12 to 24 |
| Findings and roadmap | 3 to 5 | 6 to 10 | 12 to 24 |
| Client report and handoff | 3 to 5 | 5 to 8 | 10 to 16 |
| Planning total | 20 to 32 | 39 to 65 | 84 to 160 |
These are derived planning ranges assuming an experienced practitioner, one language, cooperative client inputs, and a documented workflow. They are not industry benchmarks; replace them with your own time data after two engagements, which is the single highest-leverage pricing improvement available to you.
Loaded hourly cost means the real internal cost of the person: salary or contract cost plus employment burden plus the non-billable time their billable hours must carry. It is not the rate on your invoices. The U.S. Small Business Administration’s break-even guidance treats price, variable cost, and margin as one connected decision; this model applies that discipline to a scoped service.
Scope-first pricing calculator
Use loaded delivery cost, not a public price average. The output is a planning estimate before taxes, payment fees, sales cost, and local legal requirements.
The calculator excludes taxes, payment fees, sales cost, and local legal requirements; add those in your own commercial model. Round the output to a sensible number after the math, never instead of it.
Three worked calculations at three scopes
Worked example
Small diagnostic baseline, priced at $3,000
Declared assumptions: 16 prompts, 2 platforms, 2 runs per prompt per platform (64 observations), 5 priority pages, essential technical checks, a decision memo instead of a full report. Solo practitioner with a loaded cost of $65 per hour; $60 of tool cost; 50 percent target gross margin.
Hours: discovery 2, technical 4, prompt design 3, observation and classification 5, source map 2, findings and memo 4, handoff 2. Total 22 hours.
Delivery cost = 22 x $65 + $60 = $1,490 Price floor = $1,490 / (1 - 0.50) = $2,980, rounded to $3,000
The diagnostic exists to answer one question: does a full audit look justified? Price it low enough to be an easy yes, high enough that you are not subsidizing discovery.
Worked example
Standard client audit, priced at $9,200
Declared assumptions: 30 prompts, 3 platforms, 3 runs (270 observations), 10 priority pages, full fact ledger, competitor comparison on the same panel, client report and 90-day roadmap. Loaded cost $75 per hour for a senior practitioner; $250 in tools; 55 percent target margin.
Hours: discovery 4, technical 7, prompt design 6, observation and classification 14, source and competitor map 6, findings and roadmap 8, report and handoff 7. Total 52 hours.
Delivery cost = 52 x $75 + $250 = $4,150 Price floor = $4,150 / (1 - 0.55) = $9,222, rounded to $9,200
Note what the margin is for: it funds the sales cycle for deals that do not close, methodology maintenance, and the re-run when a platform changes an interface mid-engagement. It is not padding.
Worked example
Complex multi-market audit, priced at $25,700
Declared assumptions: two languages and regions, so two 30-prompt panels on 3 platforms with 3 runs (540 observations), 20 priority pages across two locales, regulated-industry fact review, compliance sign-off, and two stakeholder workshops. Blended loaded cost $85 per hour across a senior lead and an analyst; $600 in tools; 60 percent target margin, reflecting higher delivery risk.
Hours: discovery and stakeholders 10, technical 16, prompt design 12, observation and classification 28, source and competitor map 14, findings and roadmap 16, compliance review 6, report, workshops, and handoff 12. Total 114 hours.
Delivery cost = 114 x $85 + $600 = $10,290 Price floor = $10,290 / (1 - 0.60) = $25,725, rounded to $25,700
At this scope, price discovery separately and first. Every unresolved variable (which regions, which platforms, how many stakeholders) swings the fee by thousands, and guessing on the client’s behalf is how complex audits become unprofitable.
If a computed floor is too high for a buyer, the honest moves are to cut named scope (fewer platforms, a diagnostic instead of a full audit), improve delivery efficiency, or choose a different buyer. The dishonest move is keeping the price and quietly cutting the evidence standard, which produces the screenshot-deck audits this market is already drowning in.
Keep audit, implementation, and monitoring priced apart
The audit diagnoses and prioritizes. Implementation changes facts, content, technical controls, and profiles. Monitoring re-runs a declared panel on a schedule and reviews first-party signals. These have different owners, different acceptance criteria, and different margins, and bundling them into one undefined “GEO retainer” makes both the margin and the client’s expectations unmanageable. Sell the audit fixed-fee against the scope in the proposal template, and let monitoring or implementation be its own decision after the roadmap exists. The transition conversation is easier, not harder, when the audit was priced and delivered as a bounded product, and it is the natural expansion path described in how to sell AI visibility audits.
Mid-engagement, the fee changes only through scope: added platforms, languages, pages, competitors, prompts, runs, or deliverables, priced from the same model before the work starts. Lower the fee only by removing something named. A discount with unchanged scope is a silent decision to cut evidence quality later.
Underpricing is how audits stop being audits
A fee is only defensible if the protocol behind it isThe calculator prices the scope. This one grades the method inside it across 47 weighted checks, so you find out which sections are thin before a buyer asks what their money bought.
Score what you are quotingEstimate your observation count from the prompt-set method, run your own numbers through the calculator above, and track actual hours by workstream from the first engagement, so your second price is built on your evidence instead of this model’s assumptions.
Frequently asked questions
How much does an AI visibility audit cost?
There is no reliable public price sheet for this service. Under the declared assumptions worked in this article, the cost model produces roughly $3,000 for a small diagnostic, $9,200 for a standard agency audit, and $25,700 for a complex multi-market engagement. Those are derived planning figures from stated inputs, not observed market benchmarks, so replace the hour and cost assumptions with your own time data after two engagements.
What formula should you use to price a GEO audit?
Price from scope, not from a competitor's number: delivery hours times loaded hourly cost, plus direct tool cost, divided by one minus the target gross margin. That computes an honest cost floor, while the client's decision value and willingness to pay set the ceiling. Loaded hourly cost means the person's real internal cost, including employment burden and non-billable time, not the rate printed on your invoices.
What drives the cost of an AI visibility audit?
Three drivers move the fee. Observation volume multiplies fastest: prompts times platforms times repeated runs, so a 30-prompt panel on three platforms with three runs is 270 observations. Evidence standard is second, because saving exact prompts, full answers, classifications, and fact-ledger comparisons costs two to three times the hours of a screenshot deck. Scope surface is third: pages, competitors, languages, locations, and stakeholders each add fixed chunks of work.
Should an audit include implementation and monitoring?
No. The audit diagnoses and prioritizes, implementation changes facts, content, and technical controls, and monitoring re-runs a declared panel on a schedule. Each has different owners, acceptance criteria, and margins, so bundling them into one undefined retainer makes the margin and the client's expectations unmanageable. Sell the audit fixed fee against named scope, and let implementation or monitoring be its own decision after the roadmap exists.