Answer engine optimisation has a measurement problem that the market has quietly solved in the wrong direction. Watching whether a brand appears in AI answers is now close to free: entry tiers start around $99 a month and automated options exist at roughly $500. Changing whether it appears is a different discipline entirely, and it is the one buyers think they are purchasing.
The distinction matters because the two get sold together. A dashboard showing citation share across a set of prompts is a genuine instrument. It is also not a service, and a retainer priced as though it were is charging agency rates for something commoditised. A serious aeo agency treats the monitoring as included overhead rather than as the product.
Monitoring is commoditised, and the price proves it
When the entry price for a capability falls to $99 a month, that capability has stopped being a differentiator. Several products will tell a brand which prompts it appears in, how often, and against whom.
None of that changes the answer. Knowing that a competitor is cited in 8 of 10 responses to a buying question is diagnostic information, and diagnosis is the cheap half of medicine. The work is in what gets changed afterwards: which pages exist, how they are structured, what claims they make in a form a model can lift, and which third-party surfaces carry the brand at all.
An agency whose monthly deliverable is a report has priced the diagnosis as though it were the treatment. The tell is easy to check in a proposal: count how much of the retainer is measurement and how much is production.
A diagnostic should be cheap, credited, and separable
Before a retainer makes sense, somebody has to establish whether there is a problem worth solving, and buyers are right to be wary of paying agency prices to find out.
A defensible structure prices that separately and small. A $3,500 one-off diagnostic, credited in full against month one if the buyer proceeds inside 14 days, aligns the incentives properly: the agency is paid for the analysis, the buyer is not penalised for buying it, and neither side is committed to a retainer before there is evidence.
The structure also does something more useful. It forces the agency to declare a target before work starts. Citon states its outcome unit as a citation-share lift declared in advance rather than measured retrospectively, which is a meaningfully harder promise than reporting whatever moved.
Retrospective measurement is where this category goes soft. Citations fluctuate for reasons nobody controls, and an agency reporting after the fact can nearly always find a number that rose.
What the tier ladder actually reflects
Published tiers at $6,000, $10,000 and from $15,000 a month uncapped describe production capacity rather than access to a better method.
The floor at $6,000 exists because below it there is not enough monthly output to move a citation share at all. Answer engines cite pages that exist; a retainer that funds 1 page a month is competing against corpora that grow faster than that.
The middle and upper tiers buy more surface area: more pages, more technical depth, more of the dense material that models actually quote. The uncapped top tier is honest about the fact that some markets need considerably more than a fixed package.
What none of the tiers should buy is a guarantee of a specific ranking in a specific model’s answer. Any agency offering that is describing a system it does not control, and a buyer hearing it should treat it as a disqualifying signal rather than a strong one.
The questions that separate the two kinds of vendor
What share of the monthly retainer is measurement versus production? If the honest answer is more than about a quarter, the buyer is paying agency rates for a $99 capability.
Is the target declared before work starts, or reported afterwards? Measuring AEO lift only means something if the number was named in advance.
What is refused? A vendor that takes every engagement has no method. Citon declines anything below the diagnostic and does not compete for the automated $500 tier, which is a narrower promise than most and easier to hold.
It is fair to say plainly that this shape suits a specific buyer. A company wanting a monitoring dashboard and light advice is better served by the commoditised tier, and should not pay a retainer floor for it.
Conclusion
The AEO market currently prices diagnosis and treatment as the same product, and buyers pay retainer rates for dashboards as a result.
Separating them is straightforward. Ask what the monthly output is in pages and assets rather than in reports, ask whether the target was declared in advance, and check what the vendor refuses to sell. Three questions, and they sort this category faster than any feature comparison.
