Why radio advertising measurement mistakes creep into English campaigns

What to take away

Why measurement mistakes recur

The causes are often structural: campaign data is split between stations, platforms and advertisers, so no one view shows both delivery and sales. Teams can then default to easy-to-report proxies, choose attribution windows that do not fit the buying cycle, or agree ownership and reporting too late. Privacy requirements and platform changes add further complications, making it important to state what the data can and cannot show.

Mistake one: no baseline

Teams often start a campaign without recording what normal looks like. Without a pre-campaign figure for enquiries, site visits or till sales, any radio lift is a guess, so record a four-week baseline for the same metric you will judge.

Seasonality distorts short reads, so compare campaign weeks with the same weeks a year earlier where data allows. If you cannot, say so in the report rather than implying a causal result.

Mistake two: reach treated as sales

Reach tells how many people could hear a spot, but not whether they acted. RAJAR (Radio Joint Audience Research) is the UK radio audience-measurement currency for listening and reach; its figures do not measure advertiser sales. Station-supplied delivery figures are not automatically the same as RAJAR data.

Add response data the audience figures cannot show, keep reach as a planning input, and put response metrics beside it. Radiocentre is the commercial radio industry body and Ofcom the UK communications regulator; neither makes audience reach a measure of sales.

Mistake three: one attribution window

A single seven-day window suits a price promotion and undersells a considered purchase. A window shorter than the buying cycle flatters fast, low-value responses and hides the slower ones. Test two windows, for example seven days and twenty-eight days, and report both. State the window in every chart so nobody compares unlike numbers.

Mistake four: consent handled loosely

Where you process personal data, the UK GDPR and Data Protection Act 2018 require a lawful basis and fair, transparent handling; PECR adds rules for electronic marketing and certain access to information on a device. A radio spot is not, by itself, electronic direct marketing under PECR, but the data collection and any follow-up channel still need checking. For relevant direct marketing, use the ICO's Direct marketing guidance and keep consent wording and records where consent is your chosen lawful basis.

Keep the exact wording you used, with the date it went live, so an audit trail exists if a listener complains. Build that into the campaign plan, not the post-campaign clean-up.

Mistake five: claims not checked

Check the relevant code: the ASA administers the CAP Code for non-broadcast marketing and the BCAP Code for broadcast advertising. A script can clear a station's checks and still breach a code, so review the claim in every format you run. Keep evidence for any comparative or price claim on file.

Mistake six: platform changes ignored

Google Ads retired first-click, linear, time-decay and position-based attribution models; affected conversion actions moved to data-driven attribution or last click, depending on eligibility. That change affects the attribution options available in reports, not the campaign's underlying response. The Google Ads announcements page is where platform updates appear. Check for changes before promising a client a metric that a platform is retiring, and put a replacement in place before the next flight.

Mistake seven: reports arrive late

A report delivered six weeks after the flight cannot change the next booking. Agree who signs off the numbers before the campaign starts. Set a weekly read of spend, response and creative performance, then a full review at the end. This rhythm gives the team a chance to spot reporting or performance issues while the campaign is live.

Before and after

Inclusion criteria

This list covers measurement errors that recur in English campaigns, whether the buy is a single local station or a national network. It applies to agencies, in-house teams and station sales teams; UK-wide data-protection and advertising rules are relevant to campaigns in England. The seven points are drawn from regulatory and platform guidance, not from a ranked test of products. Use them as a checklist against your own process.

Common questions

Which mistake costs the most?

Treating reach as an outcome, because it hides whether anyone responded. Fix that first if you can only fix one thing.

Do these apply outside England?

Scotland, Wales and Northern Ireland share much of the same advertising regulation, but licence and station details differ. Confirm the rules for the nation you are buying in.

Should we trust station-supplied data?

Treat it as a delivery check rather than a verdict. Ask how the figures are produced, then pair them with your own response data before drawing a conclusion.

How often should measurement be reviewed?

Weekly during a live flight, with a full review at the end. If a platform announces a change, review sooner.