Why talent agencies review methodology matters before you shortlist

What to take away

The ASA register of non-compliant online advertisers is free to check, and it names advertisers the regulator has ruled against. Log the date of every check, starting 11 September 2026.
A defensible review scores only what you can evidence from documents, registers or written answers. Gut feel and one-off demos are not scored.
Vendor claims about audience size, roster and growth sit in a claims column until a second source confirms them.
Editorial judgement is allowed, but it is flagged and kept out of the evidence score.
Hands-on testing is out of scope. This method reads published material, regulatory records and written responses.

What does a talent agencies review methodology actually score?

Four evidence tiers, each weighted differently.

Tier one is regulatory and corporate records: Companies House filings and the ASA register of non-compliant online advertisers.

Tier two is the agency's own written material, from rate cards to case studies. Tier three is independent coverage, such as the Rebel 50 list of challenger agencies published by The Drum. That list helps separate genuinely independent creator and talent agencies from network sub-brands. Tier four is buyer feedback, treated as weak evidence because it cannot be verified.

Nothing else is scored. Office visits, follower counts and awards entries stay in the context column.

How do you separate vendor claims from independent evidence?

Every fact sits in one of two columns. The claims column holds what the agency says about itself, with the date. The evidence column holds what a third party confirms. If an agency claims a roster of forty creators and only twelve appear on its own site, log the claim and note the gap.

Brand safety needs its own check. A brand named as a client says nothing about that brand's ad compliance, so look it up separately on the ASA register.

For context on how advertising bodies define standards, IAB UK's story and remit explains the trade body's role. It helps when an agency cites "industry standard" practice without naming the standard.

Which parts of the review are editorial judgement?

Judgement covers fit, clarity and risk. Examples: whether a commission structure is easy to explain to a creator, or whether a termination clause is unusually one-sided. These are opinions, so the scoring sheet flags them with a J. A reader can then discount them.

Judgement never inflates an evidence score. A section with no evidence stays empty rather than being filled with a plausible guess.

This mirrors the discipline in talent agencies reviews and comparison methods for 2027, where scoring weights are set before any agency is examined.

What was not assessed, and why that matters

We did not run campaigns with any agency, interview creators under contract, or test software. No mystery shopping took place. Those methods raise consent and confidentiality problems.

A review built on documents and registers is harder to mislead. It cannot tell you how an agency behaves in a crisis, so keep that limit beside the score.

How should you document the review?

Documentation is what makes a shortlist defensible six months later. Use one sheet per agency, and date every entry.

Agency legal name and Companies House number.
Date the ASA register was checked.
Source and date for every roster or audience claim.
Whether each claim is vendor-stated or independently confirmed.
Judgement items, flagged as judgement.
Open questions sent to the agency, with the reply date.
Decision, with the reason in one sentence.

Contract terms and creator consent need a separate pass. The talent agencies selection checklist in England covers both before you sign.

Common questions

Is a published ranking enough to shortlist an agency?

No. A ranking tells you who someone else rated and how. Treat it as tier three evidence, then check records and written terms yourself.

How often should the review be refreshed?

At least annually, and straight after any regulatory action or ownership change. Registers and agency pages change without notice, so an undated review is close to useless.

Can a small agency score well without a long client list?

Yes. The method rewards verifiable structure: clear contracts, named contacts and consistent disclosures. A short roster with clean paperwork can beat a long one that cannot be verified.

What if an agency refuses to answer written questions?

Record the refusal and the date. It is evidence in itself, and it usually belongs in the risk column.