Media buying agencies benchmark research: a UK method

What to take away

Start with the decision, not the number

Fix the decision, the medium and the metric first. Then pin down the population, the geography, the period and the cost scope. Records that differ on any material field stay apart, unless you can defend a specific adjustment.

Ask what changes if your new result sits above or below the reference. If nothing changes, the figure is decoration. That test matters most for commitments you cannot reverse, such as a full-year television booking.

Population and geography follow from the audience work behind the brief. Set that out with audience research for a media agency brief before you collect any figures.

Geography flags for England-only campaigns

Great Britain covers England, Scotland and Wales. The UK adds Northern Ireland. PAMCo and Route publish on a Great Britain basis for the media they measure.

For an England-only campaign, flag the coverage difference in the ledger. Do not relabel a Great Britain or UK figure as England. Record instead that the figure is directional for England, and why.

Go to the currency for each medium

The IAB measurement hub routes readers to digital measurement material and industry resources. Treat it as a starting point, not one universal benchmark. Follow the exact standard behind any figure.

Public methodology pages and release notes are free to read. Subscribed planning datasets and their permitted uses are checked with the provider. Tool defaults such as Keyword Planner volumes sit elsewhere, and media buying tools on the record explains why they are not a currency.

The digital counterpart is Ipsos iris, the UK's online audience measurement service for PCs, smartphones and tablets, endorsed by UKOM, which sets the UK standard for online audience measurement. Its figures follow their own published definitions and the detailed data sits behind a subscription, so record the edition and the access route as you would for Barb or RAJAR.

Agency-side benchmarks: sources and measures

The currencies above benchmark media audiences, not agencies. Benchmarks about the agencies themselves come from other publishers. The IPA's Agency Census reports employment across member agencies and its Bellwether Report tracks marketing budget trends. ISBA publishes the media agency framework and contract guidance that advertisers use in pitches and reviews, and the Advertising Association publishes UK advertising spend data.

If the question is about a named agency, start with its registration and filings at Companies House. GroupM, Publicis Media, Omnicom Media Group and Dentsu all trade through UK companies whose accounts are on the record, which gives you headcount and financial context rather than a rate card.

Agency-side comparison turns on measures that are rarely published in full: the remuneration basis (commission, fee or retainer), pitch and transition costs, the staff cost ratio behind a plan, and the media cost index used to judge buying. There is no public table of typical fees or commissions, because those terms are agreed client by client; treat any quoted average as a claim that needs a named source, a scope and a date. Media auditors and advisers such as Ebiquity and MediaSense hold cost and performance comparisons through commissioned work, so ask for the publisher, the edition and the period before a figure goes into the ledger.

Build a comparison ledger

Give each candidate figure one row. Record the publisher and the exact page or dataset edition, with the access date. Add the medium, population, geography and period.

The row also needs the metric name, its unit, its base, any weighting and any exclusions. State whether the value is observed, estimated, modelled or forecast. Note the proposed use, the limitation and the approver.

Save the original extract under your organisation's data rules. The ICO UK GDPR guidance is the reference point when personal data sits behind the figure. A percentage without its base, or a cost without its inclusions, is not ready to compare.

Example ledger row

Reject false matches early

Do not compare planned reach with delivered reach. Keep people apart from devices, and gross contacts apart from unique audience. Platform-attributed actions do not belong beside client-accepted orders.

Separate Great Britain, UK and England records. Dates matter too: a seasonal campaign and an annual average may describe different conditions even when the metric definition matches.

Response figures depend on the attribution rule. Read choosing a media attribution method before treating two response benchmarks as compatible.

If a source publishes a range, preserve it. If it publishes no suitable figure, record the gap. There is no sound basis for inventing a midpoint or a typical agency rate.

Turn the benchmark into a decision

Where the choice is material, run sensitivity checks with more than one compatible reference and explain why each belongs. A short verdict beats a polished chart built from mismatched samples.

Common questions

Where can I find actual UK benchmark figures?

Start with the published material from each currency body: Barb, RAJAR, PAMCo and Route. Public methodology pages and release notes are free; detailed planning datasets sit behind a subscription or licence.

What should I do if no source publishes a suitable figure?

Record the gap in the ledger and say that no compatible figure was found. Do not invent a midpoint, a typical agency rate or an England adjustment to fill the row.

Why should I not blend two figures with similar labels?

Mismatched samples describe different populations, periods or cost scopes. Keep them apart, or explain the adjustment, because compatibility comes before any search for an impressive average.

How do I check the provenance of a quoted figure?

Ask the agency for the publisher, the edition and the access date. If you are assessing the business behind the quote, its registration, filings and directors are on the record at Companies House.