How we compare Facebook ad agencies: the scoring explained

What the score is trying to answer

The question is not 'which agency is best'. It is 'which agency can a tradesperson verify, afford and leave'. A roofer choosing a Facebook ads agency has no way to audit campaign skill in advance. What can be assessed from the outside is disclosure, fit, commercial terms and evidence, so that is what we score.

Each agency is marked out of 10 on seven criteria. The criteria are weighted, the weighted marks are summed, and the total is scaled back to a score out of 100. A score above 70 means we found little to query. A score below 50 means the burden of proof sits entirely on the sales call.

The seven criteria and their weights

Proof and fit carry the most weight because they are the two things most likely to predict whether the engagement works. Disclosure carries the least because a missing founder name is a yellow flag, not a red one.

Published proof: what a 4 and an 8 look like

A 4 is a site with testimonials but no figures, or figures with no trade or timeframe attached. Most agencies in our database sit here. Rated Media and Scale Through Us are examples: generic testimonials, no case studies.

An 8 is a library of case studies with numbers across several trades, as One Base Media publishes, or anonymised results with a clear metric, as Roofing SEO Services does for map pack rankings. A 10 would add independently verifiable client references on the page, which no agency in our set currently offers.

We do not treat 'over 1,000 businesses helped' or '300+ trade companies' as proof. Those are scale claims. They count under disclosure only if the agency shows how the number was reached.

Trade fit: naming is not evidencing

An agency earns fit marks for a trade in two steps. Naming the trade on its site earns up to 5. Publishing a case study or named client in that trade earns the rest.

This is why a broad agency can score lower for a specific trade than a narrow one. One Base Media names 15 trades but a locksmith is not among them, so for our locksmith comparison it scores on general trades experience only. Landscape Marketers names only four trades, but has case studies in three of them.

Meta depth

This site is about Facebook ad agencies, so an agency that sells Meta as a core product outscores one that lists 'social media marketing' among eight services. Lead Wave, which is built on Facebook ads, scores high here. Trustmark Trades, which runs no paid ads at all, scores 1, and we say so rather than excluding it, because readers comparing options deserve to see the search-only alternative.

Depth is judged on how the agency describes its own work: whether Meta leads its service list, whether it publishes Meta-specific results, and whether creative production is mentioned. A Meta agency that outsources all creative has a weaker claim to depth, a point covered in creative production: in-house or agency.

Price transparency and commercial terms

Transparency is scored on what is public. Published tiers earn 8 or more, as Roofing SEO Services (£499 to £2,599 a month) and Leads 4 Local (£349 to £999) do. A pricing page that exists but shows no figures earns 4. Nothing at all earns 2. We do not penalise an agency for being expensive, only for hiding the number.

Commercial terms are scored separately. Rolling monthly with no setup fee is the benchmark and earns 8. Twelve-month minimums lose marks. Exclusivity promises neither add nor subtract; we treat them as neutral because, as argued in exclusivity is mostly a pricing trick, they usually fund a price premium rather than a benefit.

Asset ownership

The single question is: if you cancel tomorrow, what do you keep? An agency that runs ads from your own Business Manager, builds the pixel on your domain and hands over the website earns full marks. One that runs everything from its own accounts and leases you a site earns 2.

Where a site does not say, we score 4 and flag it as a question to ask. The full reasoning is in who owns the ad account.

What we deliberately do not score

• Awards and partner badges. Google Partner status reflects spend thresholds and certifications, not client results.
• Social media follower counts or YouTube subscriber numbers.
• Design quality of the agency's own website, beyond whether it discloses the facts above.
• Promised cost per lead. An agency cannot know your cost per lead before running your campaign.
• Our own experience as a client. We have not engaged any of these agencies and do not claim to have.

A worked example

Take a hypothetical roofing agency that publishes three case studies with figures (proof 7), names roofers and shows roofing results (fit 9), sells Meta as its main product (depth 8), publishes a from-price only (transparency 5), offers rolling monthly with a £500 setup fee (terms 6), runs ads from its own account (ownership 3) and names its founder and town (disclosure 8).

Weighted: 7×0.20 + 9×0.20 + 8×0.15 + 5×0.15 + 6×0.15 + 3×0.10 + 8×0.05 = 6.75 out of 10, or 68 out of 100. A good agency pulled down mainly by the ownership question. Fix that one clause in the contract and it moves into the 70s. That is the point of the scorecard: it shows you what to negotiate, not just who to call.

Frequently asked questions