Commercial radio’s ability to embrace technology is fuelling growth, writes Global’s Faye McDowall and Mark Hatwell.

The first commercial radio station launched in the UK in October 1973. And, 50 years on, commercial radio is stronger than ever, now attracting a record weekly audience of 39 million people and revenues at an all-time high of £740m in 2022.

At the heart of radio is the power of human connection. Familiar voices, funny stories and great music that makes everyone’s day brighter.

Because radio has always been about people and real human relationships, it’s adapted and evolved alongside changes in consumer behaviour. It has also embraced innovation in technology, continuing to change what’s possible for audiences, broadcasters and advertisers.

Radio is no longer limited to the constraints of a single ‘device’, for example. It has dozens of entry points. And each of these offers advertisers creative ways to connect with audiences at scale.

In 2014, Global launched DAX, a pioneering digital advertising exchange which connects brands with audiences at scale across audio and outdoor. This gave advertisers access to audio content in one place, as well as increased sophistication in targeting and measurement. And these metrics underpin the effectiveness of the channel. According to the latest data from Dentsu, radio is 10x more efficient than online video ads.

Digital audio has entered a new era, combining emotional impact with precision targeting and measurement, made possible by innovative technology.

Alexa, give me audience insights

It’s only been eight years since Amazon introduced Alexa to the UK, sparking one of the most successful consumer technology launches we’ve seen. Smart speakers are now active in 60% of all homes, reaching over 13.8 million people.

Radio is the most popular use for these devices. Smart speakers account for 17% of all commercial radio hours. And this figure is only growing. But although smart speakers are experiencing huge numbers, these streams have sat behind the walled gardens of the tech giants who own them and haven’t been able to provide the digital identifiers that marketers need.

Advertisers can’t see exactly who engages with their ad campaigns through smart speakers. But they don’t need to know who they are exactly. They just need to know what they’re like.

People aren’t as unique as we often think. Global’s data scientists have trained AI to read clues about audiences and filter unidentified listeners into targetable, “look-a-like” behavioural segments.

Global’s AI models can analyse audio consumption from identifiable users and look for similar habits in their unidentifiable counterparts. Our data scientists currently build audience segments from around seven million data points every week.

For example, we can take what we know about our identifiable audiences – the streaming services that listeners subscribe to, articles they have read, videos they have watched, and other listening behaviours. We might know if they’ve read about the new Marvel release, watched interviews with Barbie stars, or frequently watch Disney+. From this we can see the overlaps in listening behaviour and build our segments from identifiable users and those exhibiting similar behaviours from the non-identifiable group.

AI is only as good as the data it is trained on, though, so we asked Accenture to evaluate these AI audience models based on their criteria ranging from rigour in modelling, consistency in approach and mitigation for any potential issues.

Making audience targeting on smart speakers smart

The AI-powered tech that fuels DAX has made the untargetable, targetable.

In fact, across the whole of DAX audio, the targetable audience now available to advertisers has increased by a factor of ten since the adoption of this technology. And, more than this, it helps brands and advertisers comply with privacy laws as the third-party cookie crumbles.

We’ve seen this approach drive better performance results for clients. For one in particular, during a recent telecoms campaign targeting small business owners, a 24% increase in engagement amongst DAX audiences compared to standard audiences was observed. 

As we continue to improve, mature and future-proof these data sets we expect to understand user behaviour even better – with oversight of more content consumed by an individual, as well as their location and age and gender. 

Both AI and radio have changed a lot over the last fifty years. But their core properties have not. AI is a powerful driver of efficiency and calculation – and radio is an unrivalled medium for harnessing human connections. In the next fifty years, these qualities will become more interwoven, creating better experiences for advertisers and audiences alike.