How to choose an answer engine optimisation agency.

Five questions, and our own answers to them.

These are the questions that separate a diagnosis from a subscription. We have answered them for ourselves below, including the two we answer worst.

1. Show me a client's visibility before and after, with the actual prompts.

Our answer today: we cannot. No client of ours has agreed to be named, and we do not publish a client result without permission. That is a real gap, and it is the single most common reason a buyer picks somebody else over us. What we publish instead is the method in full, a sample report whose figures are openly invented to show the shape of a finding, and a pre-registered study protocol with the conditions that would prove us wrong written down before the run. If a published client result is what decides your purchase, say so on the call and we will tell you plainly that someone else has one and we do not.

2. Can you show citations from each engine separately?

Yes, and we refuse to do it any other way. Every engine is reported on its own denominator and they are never averaged into one score. A brand can sit at a third on one surface and at nothing on another, and a blended number hides both facts. The counting rules are published, including the four tiers that keep a passing mention out of a recommendation rate.

3. Do you work on Reddit, Quora, third-party publications and review platforms?

No. We assess that layer and we do not sell it. This is the honest limit of what we do, and the answer that should decide whether we are the right studio for you.

Third-party authority is real and it matters. What an agency can actually do about it is place content on sites it does not own, which takes months, costs money separately, and is a different trade from the one we practise. We measure where you stand on it, name who holds it in your category and on the strength of which pages, and set out what would move it. Then you either run that yourself or hire someone whose job it is. If buying placements is the main thing you want, hire an agency that does it. Several below do.

The reason we draw the line there is published on the site already: the rented and owned split. A placement you bought on someone else's page is visibility you are renting, and it lasts exactly as long as that page ranks. We would rather build the half you keep and tell you honestly what the other half is worth.

4. Will you create content, fix technical SEO and build brand authority?

The first two, yes. The third, no, for the reason above. A build is an agreed scope, quoted up front: entity resolution, structured data matched to real answers, answer pages for the money questions, and the feed, indexing and crawlability work that decides whether any of it is readable. Typical scope is around ten answer pages plus schema across your main templates, agreed in writing before anything starts. It is built on your domain, published in your name, and it is yours whether or not you keep working with us.

5. Can you show referral traffic and conversions, not just visibility percentages?

Partly, and we will not pretend otherwise. Generative referral traffic that already exists is read from your own Search Console, as first-party evidence that the surface sends anyone at all. A naming rate is not a demand figure, and no finding we publish traces a mention in an answer to a sale. Nobody in this market can honestly close that gap today, and an agency that says it can is modelling, not measuring.

What each studio actually publishes.

Read on each company's own site on 4 November 2026. "No published price" means the site asks you to book a call for a quote, not that the work is cheap or expensive. A capability listed on a website has not been independently verified by anyone, including us.

Sources, each read on 4 November 2026: unveilrai.com · prohed.com · infidigit.com · ibsfulcro.com. Figures attributed to another company are that company's own published claims, not measurements of ours.

One correction worth making, because it is about us. A widely-shared AI-generated comparison of this market quotes a monthly price for two of these agencies. Neither price is published on either company's site. We mention it only because the same answer priced us accurately, from our own published page, and then placed us as the most expensive monthly option against a figure that does not exist. If you are comparing on price, compare published prices, and ask the others for theirs.

Where somebody else is the better answer.

Hire a managed AEO service if you want one supplier for everything, including third-party placements, and you would rather buy an outcome than a diagnosis. Unveilr publishes a dated D2C case study with before and after figures and a linked methodology, which is more than we can show today. If a demonstrated client result is what decides it for you, that is a reasonable basis to decide on.

Hire a performance agency if AEO has to connect to paid media and conversion rate, and the person signing it off wants AI visibility inside one revenue conversation rather than beside it.

Hire an enterprise SEO operation if the site is large and the work is mostly scale: thousands of templates, many markets, a migration in flight, and an existing organic programme that AI search has to fit into rather than replace.

Hire us if you want to know exactly what is wrong before you spend anything, you want each engine reported separately with the denominator printed, you want to see what survives when live retrieval is withheld, and you want the fixes built on your own domain in your own name with no retainer to escape. The price is published, the method is published, and the audit fee comes off the build.

Do not hire anyone yet if your category has no answer-engine demand. We publish a post on when this is not worth doing, and it costs us work.

What this comparison cannot tell you.

It compares what five companies publish, which is not the same as how well any of them executes. We have not run any of their audits, seen any of their reports, or spoken to their clients. A published case study is evidence that work was done and reported; it is not an independent audit, and neither is ours would be.

Every competitor claim here is quoted or summarised from a public page on a single date, and marketing pages change. Four of the five publish no price, so the pricing row compares a number against an absence rather than against other numbers. And we are one of the five, which is a conflict of interest no amount of careful wording removes. The correct response to this page is to ask all five the same five questions and compare the answers yourself.