How to assess B2B demand generation agency reviews

What to take away

What counts as an agency review?

A review is a structured opinion about an agency's suitability for a defined buyer. It is not an audit and not a performance test. The reviewer reads what the agency publishes, checks what third parties record, and scores against stated criteria.

That definition matters commercially. Two agencies can score differently on the same evidence base because the weighting differs. So the review should publish its weighting.

Scope is part of the definition. A review should say which buyer it serves, which sectors it covers and how many agencies it examined, so you can judge whether it applies to you.

For a wider view of how review and comparison methods fit together, the review and comparison methods guide sets out the scoring frames used across this section.

What evidence is assessed and what is not?

Assessed: registration details, filed accounts, published case studies, contract templates where shared, complaint records held by regulators, and the agency's own statements about method and measurement.

Not assessed: campaign performance inside a client account, pitch behaviour, staff conduct in private meetings, or the accuracy of a client's own reporting. No reviewer on this publication has run a paid engagement with the agencies discussed.

Independent evidence carries more weight than vendor claims. An agency's claim of a 40 percent pipeline lift is a claim. A filed account showing revenue concentration in one client is independent evidence.

Dates matter as much as sources. A filed account from four years ago says less about an agency now than one filed last year, so a review should record when each piece of evidence was published.

How should buyers read a scorecard?

The table below shows a four-part scoring frame. Percentages are illustrative examples of weighting, not a published standard.

A buyer weighting data ownership at 40 percent would produce a different ranking from one weighting sector depth first. Neither is wrong. The published weighting is what makes the review usable, and it should be stated alongside any agency pricing comparison.

Weightings are not neutral. If you sell a complex product with a long buying cycle, method disclosure and evidence quality usually deserve more weight than sector familiarity.

The guide to product comparisons is worth reading next if your shortlist mixes agencies with software vendors, because the evidence standards differ.

Which external sources support the assessment?

For buyer ratings and profiles, check Clutch; G2 is useful for software-provider reviews when a shortlist includes tools. For editorial coverage of the sector, consult The Drum and B2B Marketing. Treat these as sources to inspect, not independent proof of an agency's performance.

For market context, the Office for National Statistics' E-commerce and ICT activity, UK release covers UK businesses' use of digital technology. It can provide context, but it is not a direct measure of demand generation spend.

The Data (Use and Access) Act 2025 amends the UK's data-protection framework, including the UK GDPR and the Data Protection Act 2018. The Information Commissioner's Office (ICO) regulates data protection in the UK. Lead-generation activity may also engage the Privacy and Electronic Communications Regulations 2003 (PECR), particularly its rules on direct marketing.

GOV.UK's legislation and regulation guidance explains where to check which rules apply to a business, and it is a sensible starting point before accepting an agency's compliance claims.

Together, ONS data, Companies House filings, published ICO records and the relevant legislation provide context for a review. Check them before accepting an agency's own account of its market or compliance.

Common questions

Do reviews verify agency performance claims?

No. A review can check whether a claim is published, dated and specific. It cannot verify that the claimed result happened inside a client account. Treat unsourced percentages as marketing.

How many agencies should a review cover?

Three to five is workable for a shortlist. Fewer gives no comparison; more turns the exercise into a reading task rather than a decision.

What should trigger a second round of checks?

Any claim that cannot be traced to a named client, a filed record or a regulator. Also check whether the agency will commit to data ownership and exit terms in writing.

Are scores comparable between reviews?

Only if the weighting is published and unchanged. Compare the criteria first, then the scores. A score without its weighting tells you very little.