Audio ad formats UK listeners actually respond to

What to take away

Why format changes response

Listeners process audio differently depending on why they tuned in. A commute, a workout and a kitchen clean-up create different attention levels. A host-read mention can feel like part of the programme, while a produced spot signals a break.

That difference affects recall. A familiar voice can carry a claim with less resistance, but it can also blur the line between editorial and advertising if the host is not briefed well. Produced spots are clearly ads, which can help with disclosure and repetition.

Speech radio and podcasts create a leaning-in environment. Music radio often sits in the background. The same script can therefore perform differently on LBC and on a music station such as Capital, even with the same audience size.

RadioCentre's published audience research can guide station choice, but it does not tell you how your creative will perform.

The CAP Code requires ads to be identifiable, so host-read formats still need a clear ad marker where the rules require one (ASA advertising codes).

Host-read vs announcer-read: where each works

In the UK, host-reads are standard on commercial speech stations such as LBC and talkSPORT and on podcasts sold through Acast and Spotify. Announcer-reads carry most of the inventory on music networks such as Global's Heart, Capital and Classic FM and Bauer's Magic and Kiss.

A host-read is not automatically better. If the presenter rushes the copy or sounds unconvinced, the ad can lose impact. Announcer-reads give you more control over pace, pronunciation and legal wording.

A mix of the two lets you compare response without changing every other element.

If you need help shaping the copy, our guide to writing a radio script UK listeners will act on covers structure and calls to action.

How each format is bought. Radio spots are sold as cost per thousand (CPT) against a station's published RAJAR audience. The same 30-second script therefore costs more on a large London station than on a smaller regional one, and announcer-reads usually come from a rate card with volume discounts across a network.

Podcast host-reads are commonly a flat fee or a cost per thousand downloads, and the presenter's time sits inside that price. Co-produced spots add studio time and approval rounds. Marketplaces such as Global's DAX let you buy across stations and podcast shows to a fixed budget.

Matching length to the job

Most UK radio campaigns still use 30-second spots because they allow a full proposition and a call to action. Fifteen-second spots suit frequency and reminders. Ten-second spots can work as tags around longer activity.

Podcast host-reads often run 60 seconds or more because the presenter needs room to sound natural. That length can improve recall if the host keeps the message focused. If the read drifts, the brand gets lost.

Use a 30-second master and cut a 15-second version for stations that sell shorter breaks. The shorter cut still needs the brand and the next step to land clearly.

This approach does not prove causation on its own. It gives you a practical signal about which format fits your audience and message.

Measuring recall by format without fooling yourself

Recall is not the same as response. A listener may remember a funny host-read but still not visit the site. Pair recall questions with a measurable action, such as a branded search, code or landing page.

RAJAR publishes UK radio reach and hours every quarter, and its figures put weekly radio listening at around 49 million adults. It does not report ad recall, so there is no public UK recall figure broken down by format. The benchmark that matters is therefore your own, matched on format, length and audience.

Attribution that links a listener to a code or landing page usually involves personal data, so the ICO's direct marketing guidance sets the consent rules for measuring it.

For podcast tracking, the UK GDPR statute page covers the lawful basis and consent wording you need.

Use the same measurement window for every format. A host-read may take longer to bed in than a produced spot, so compare like with like. Keep a simple record of format, length, placement and response.

To build a fuller attribution plan, our guide to podcast response tracking sets out the steps.

Common questions

Does host-read always beat announcer-read?

No. The answer depends on the show, the offer and the audience.

What audio ad length works best in the UK?

There is no single best length. Thirty seconds remains common on radio, while podcast host-reads often need 45 to 60 seconds to sound natural.

How do I test audio ad recall by format?

Compare branded search, code use or landing page visits alongside a short recall question. Keep the test running long enough to avoid one-off noise.