How we verify

1. Who gets listed at all

We list every kind of agency. What’s strict is where a firm appears: one category, the one it’s primarily focused on. On a general directory the specialist and the everything-shop sit in the same list, so the specialist disappears — which is why the first thing this site does is turn most applicants away. Of the firms we’ve looked at, we list 6 of 15.

A firm can have other service lines. We publish those too. What it can’t do is be listed under something that isn’t its main thing.

There are four routes in. A firm needs one.

1. The category is 50% or more of the business

   A majority, so nothing else the firm does can be bigger. We measure it against what a firm has published about itself, not the number it emails us. A declared share gets multiplied by a credibility factor and capped near what the record supports. Where the two disagree, we show the gap instead of picking one.

2. Or it leads the next-biggest service line by 2× or more

   A firm at 45% here and 20% everywhere else is a specialist, and a bare majority test would throw it out. A firm at 34/33/33 isn’t, and “largest wins” would admit it and call it one. Twice the runner-up separates those two cases with one number — and it’s the hardest part of this rule to game, because a firm can’t claim a 2× lead over its own second service line without that being visibly untrue on its own website.

3. Or 2 of 5 specialisation signals, each with a public URL

   A flat percentage would empty several categories. Almost every good AI SEO firm started as an SEO firm; almost every automation firm started in ops. A rule that excluded the ones that rebuilt their delivery model around a category would be tidy and useless. So the floor stays where it is and this is the other door — but a signal without a link doesn’t count, which is what stops it becoming a checkbox.

   • A named, publicly documented methodology the firm authored for this category — not a blog post about someone else's.
   • Tooling the firm built and operates itself for this work, publicly demonstrable.
   • The service menu itself is this category only — no separately sold service lines outside it.
   • A publicly named measurement stack for this category, i.e. they can show what they claim to move.
   • Named individuals whose stated roles are specific to this category, not generalists with a relevant title.

4. Or the firm is native or vertical to the category

   A company whose own product is the work has no meaningful “share” to measure. An agency doing fairly ordinary marketing, but only for one market, is exactly the specialist a buyer in that market wants. Both are specialists in the way that matters and neither would survive a percentage test. Every competitor files them all under one label and loses the difference.

When a firm doesn’t clear the bar for the category it applied under, the usual reason isn’t that it’s a bad firm — it’s that something else is its main focus. We say which, and point them at that category instead. Being listed in the wrong place is worse for a firm than not being listed.

We publish who we turn away, with the figure and the sources — but only when we’ve actually measured it. “We couldn’t check” never gets published as “they failed”. That difference is the whole distance between a directory and a lawsuit.

Check yourself against this rule →

2. What we check

Two lists. One every firm faces — is this a real business, trading, with people you can name — and one written for its category alone. The AI SEO list asks things an AI PPC firm would find irrelevant, and the other way round. A shop that does a bit of everything can’t quietly pass the second list, which is the point.

1. Is AI search actually their business?

   • AI search is a named service, not a blog topic — I look for a dedicated AI search or GEO service page, not a paragraph about AI bolted onto the traditional SEO page.
   • When the AI search page first appeared — I check the archive for the first time that page existed. Conviction before 2024 reads differently from a page that went up six weeks ago.
   • The traditional-SEO vs AI-search split is knowable — I work out from their own site how much of the practice is AI search and how much is classic SEO wearing a new label.

2. Can they measure the thing they sell?

   • A named AI-visibility measurement stack — I check whether they name how they measure AI visibility at all — a tracking product or their own tooling. Anyone selling AI search visibility with no stated measurement method is selling a feeling.
   • A defined AI-search KPI — I look for a stated metric — citation share, prompt coverage, answer presence. Rankings is not an AI-search KPI.
   • A case study reporting an AI-search number — I check that at least one result is an AI-search figure rather than an organic-traffic figure. Selling AI search and reporting Google traffic is this category's signature swap.

3. Who would actually do the work?

   • A named search lead with a traceable public record — I look for a real person with published work, talks or studies attached to their name — not a team of experts.
   • Team size against client count — I compare the team page against the client wall to see whether the ratio is plausible.

4. What does an engagement cost, and how do you leave?

   • Stated pricing model — I look for a named model — retainer, project, performance or hybrid — rather than one hidden behind a call.
   • Discoverable contract length and exit terms — I check whether the commitment and how to leave are findable without signing anything first.

Five ways a check can end, and we show all five

Verified

We found it in public record.

Stated

The firm says so and we haven’t been able to confirm it.

Pending

We haven’t run this one yet. That’s our backlog, not their failing, so it stays in the denominator.

Unavailable

We tried and the source blocked us. This never shows up as a failure.

Not applicable

There’s no such source for this firm. Leaves the denominator entirely.

Treating “a bot wall blocked us” as “this agency failed” is how a verification engine turns into a defamation problem. The five labels exist so it can’t happen by accident.

3. The TF Score

It measures how much we could actually verify — not how good the agency is. A firm can be excellent and score low because we haven’t got to it yet. We say so on every page the number appears on, because a 0–100 number beside a company name gets read as a rating unless you keep saying it isn’t one.

60 points are universal and 40 are specific to AI SEO.

2 pillars are declared but not yet scored, because we have no checks behind them yet. They are removed from the denominator rather than scored zero — counting an unasked question as a failure would defame every firm equally, and dropping it silently would inflate every score equally.

What a claim is worth

Evidence ages. Each check carries its own shelf life, and a result past it is worth less than a fresh one — a verified fact from three years ago is a historical note.

Bands

• Well evidenced — 80 and above
• Evidence-backed — 65 and above
• Developing evidence — 45 and above
• Limited verified evidence — 0 and above

How the order is decided

Firms are ordered by evidence confidence, then by category delivery record, then by how recently we checked. Scoring and ranking are separate on purpose — one number can’t honestly mean both “our evidence is strong” and “look at this one first”.

Ranking is never for sale. The code that ranks firms reads a database view that can’t see payment, subscriptions, claim status, sponsored slots or reviews at all — and the deploy stops if the order changes between a paying and a non-paying state. Sponsored placement exists: it’s capped, sits above the list, is labelled on every card, and never reaches the structured data an AI assistant reads.

4. Proof tiers

Separate from the score: how well a firm’s own claims stand up. Tiers are worked out in code from what reviewers record, never set by hand — a badge someone can toggle is a badge for sale.

Publicly verified

We checked this firm's core claims against sources anyone can open, and they held up.

Confidentially reviewed

A reviewer saw evidence for this firm's core claims that can't be published — client work under NDA, usually.

Self-reported

These are the firm's own statements, published as such. We haven't been able to confirm them yet.

Insufficient public evidence

There isn't enough on the public record to say much either way. That's a statement about the evidence, not about the firm.

5. Tools

A firm can tell us what’s in its stack and we label that as its own statement. Calling a tool a specialism is a claim about capability, so it needs backing up; without it we show the tool as a stack entry and say on the profile that we walked the claim back. Naming tools can’t move a firm up any list.

6. Getting it wrong

We will. When we do, fixing it is free and open to anyone— claimed profile or not, paying or not. Charging a company to correct something untrue we published about it isn’t defensible in any country we operate in. Only the promotional right of reply sits behind a claim.

Report something wrong →

Reviews

Reviews never move a firm’s position and never unlock a badge. They’re a paid feature, so either would make both buyable by proxy. We don’t reprint competitors’ review scores either — those were collected for their platforms, not ours. Where a firm has a profile elsewhere we say so and link it, and leave the number where it was earned.