How to weigh video marketing agency reviews before you trust them

The shortlist pauses before a signature

In this hypothetical scenario, a team has a £60,000 annual budget, a shortlist of three video marketing agencies and a possible twelve-month retainer. Before choosing a supplier, it needs to decide whether the reviews and claims submitted deserve confidence, rather than pretending to know how an untested campaign would perform.

It sets a clear boundary: this is a desk assessment, not a trial of delivery. No campaign is run, no studio is visited and no staff interview is conducted. That limit is recorded in a review method, so evidence visible on paper is not mistaken for observed performance.

The documents are sorted before they persuade

The buyer puts each submission into a source category before comparing suppliers: vendor claim, independent evidence or editorial judgement. That first pass prevents polished presentation from gaining the authority of a record or outside account through repetition alone. It also makes the eventual score traceable.

Pitch decks, testimonials and award entries remain claims made by the supplier; company records and regulatory publications sit in the independent-evidence column; the reviewer's interpretation stays separate. A page with no stated method may be a lead-generation asset, rather than a dependable assessment, so its conclusions are not accepted at face value.

Each item is logged with its publisher, date, stated method and evidence category. A case study remains useful when its figures can be checked, but repeated claims are not separate confirmations if each page traces back to the same announcement. The selection checklist helps verify a supplier's baseline details.

That distinction is the hinge of the scenario: a chart can illustrate a source, but cannot independently prove its claims. The buyer keeps its own notes apart from agency material and records why each item was trusted, discounted or left unresolved, rather than letting one compelling story dictate the ranking.

The records sharpen the questions

Public records and compliance information do not prove campaign quality; they show whether the agency's published claims and basic obligations merit further scrutiny. In this scenario, the checks narrow the unknowns instead of announcing a winner or guaranteeing delivery.

The team reviews Companies House filings, accounts, directors, filing history and persons with significant control, then notes whether an account or confirmation statement is overdue. These details may support or contradict a supplier's self-description, but they do not predict results for the proposed brief.

Before accepting a growth story, the buyer consults ONS business demography as context for business statistics, market size and sector trends. This does not identify which shortlisted agency will perform well. The source and its limits are recorded together, so a broad market trend is not misrepresented as evidence about an individual supplier.

Data handling forms a separate line of enquiry. For email or SMS activity, the team reviews data protection arrangements, including consent, the soft opt-in and a lawful basis to record before signing a data processing agreement. Insurance and advertising compliance also stay on the checklist, rather than serving as proof of creative quality.

Claims in the proposed video need their own compliance check. The buyer follows ASA rulings, because rulings may affect campaign claims. Those rulings are useful context, not a complete assessment of an agency or its handling of a particular project.

Where agency reviews appear, and how to test them

The shortlist is hypothetical and this article names no agency. The tests apply to any review you find on the sources below.

Three tests apply to every review. Can you identify the reviewer and confirm they bought a comparable service? Does the review describe the work, the brief and the outcome, or only praise? Does the same reviewer or wording appear elsewhere?

The law helps too. Fake or hidden-incentive reviews are banned under the Digital Markets, Competition and Consumers Act 2024, which the CMA enforces. For the data side of any shortlisted agency, the regulator is the ICO. The law that applies is the UK GDPR and the Data Protection Act 2018 for personal data, and the Privacy and Electronic Communications Regulations 2003 (PECR) for email and SMS, where the soft opt-in sits.

The comparison produces a lead, not a verdict

After sorting the evidence, the buyer applies one weighting across the shortlist: 50% for independent material, 30% for verifiable supplier claims and 20% for editorial judgement. These weights are illustrative. This makes the decision visible and stops presentation polish from silently changing the test between agencies.

Agency A's weighted total is 6.1 out of 10 (0.5 x 8 + 0.3 x 3 + 0.2 x 6). Its relatively strong public-record score does not outweigh its weaker case-study score. The calculation shows how the evidence mix shapes the result; the same criteria can be applied to each supplier without changing the test to favour a preferred submission.

Budget fit, evidence quality, contract exit terms and data protection remain visible in the comparison. The product comparison resource also sets out fields for comparing tools and platforms. The result is still a desk assessment, not a verdict on delivery quality. For the wider process, see seven ways to run agency reviews properly.

The buyer carries forward a short set of rules

The team treats the score as a prompt for buyer checks, not permission to sign. It plans reference calls, asks about communication and invoicing when deadlines slip, and uses a paid discovery phase and written brief before deciding whether to commit to the retainer.

These rules keep the assessment proportionate to what was actually checked. The buyer can say Agency A led under the stated weighting, but cannot claim that a campaign was tested or delivery quality proven. The result is a structured starting point, followed by references and discovery work matched to the brief.