The hills are alive with the sound of fuzzy logic

Which is probably why I ended up in media rather than as a copywriter.

Had I been in media at the time of course, I would have just given her a radio diary and asked her to remember which hill she was listening to and when. Fortunately, as we know from Mary Poppins, Julie had perfect recall. I mean anyone who can say supercalifragilisticexpialidocious backwards certainly knows a thing or two about navigating the audioverse. And so even though she was a flibbertigibbet and a will-o-the wisp, we could accurately calculate her peak-time listenership, so to speak.

Unfortunately, Julie was not a typical nun and was therefore not a statistically reliable sample for projecting total peak listenership listenership at the Nonnberg Benedictine Convent in Salzburg.

So, how do you solve a problem like Maria?  

Of course, in the advertising industry, despite the evidence of our own ears, we continued with the diary method of radio audience measurement until recently. In the case of South Africa, right up until 2023 as it happens. The planned shift away from diaries was entirely consistent with global best practice, but efforts at migrating from diaries to a more holistic and inclusive form of audio listening behaviour, have not yet yielded the hoped for results.

Since 2023 of course, the BRC has been unable to measure “hill listenership”, or any other form of radio station listenership for that matter.

The absence of regular & reliable radio audience research might present a challenge when it comes to media trading and pricing, but it should never become a barrier to strategic media thinking. Sometimes we just need to view the old data from a different perspective.  

The recent release of Brand Mapp 2025 data is a case in point.

As far back as the 2019 Nielsen Digital Consumer Survey and Nielsen Fusion Study 2020, we have been able to report on radio listening collectively across audio formats and devices. But when it comes to BRC RAM Amplify data we have not been able to apply that perspective to specific stations, with any degree of statistical reliability. And, as any planner will know, data spreadsheets populated with the symbols # and * don’t exactly fill you with confidence. Certainly not at the 95% level anyhow.

The release of Brand Mapp 2025 changes the playing field because now we have a measure, at scale, of station listenership on a conventional radio device and digital radio (listening online and listening via streaming). Where BRC RAM 2023D reports only 9% of daily radio listeners listening to any form of digital radio station content,  Brand Mapp 2025 reports, in some instances, up to a 6-fold increase in that number.  

Of course at a simplistic level, the temptation is to use this data insight to engage in the media industry’s latest frivolity (called stackable reach) by simply adding all these listening elements together, to “double” the audience. Stackable reach is essentially what we have known for decades as Total Duplicated Audience.

Conventional Radio listening + Digital Radio Listening = Total Duplicated Radio Audience.

But, one person listening to a radio station, on different listening devices, doesn’t suddenly become two people. Not statistically, nor in real life. Where it comes to Julie Andrews, there can be only one!

As media strategists, we know that radio listening across analogue radios, mobile phones, car radios, and tablets involves complex, overlapping, and imprecise behavioural patterns. Consumers may switch devices throughout the day and have varying levels of listening intensity depending on location, mood, or context.

Radio listeners today, like Maria, are also flibbertigibbets.

To extract the real insight from these holistic listening filters what we need to do is to calculate unduplicated audience reach. An unduplicated audience reach refers to the total number of different people exposed to a media schedule at least once, with no double‑counting.

For example, if we examine a basket of stations, we can see that BrandMapp 2025 reports up to 65% of listeners to a station may well be tuning into that station from a platform or device which is not a traditional radio set. And that some four out of every ten digital listeners also listen to conventional radio. This is a remarkable insight into the evolution radio listening behaviour.

But if want to extract maximum value from BrandMapp 2025 we need to go even further than this level of reporting, because BrandMapp listenership is not a currency. We need to apply a bit of fuzzy logic to measure the implication of this listening pattern for a station’s currency listenership.

Fuzzy logic is a mathematical approach designed to handle uncertainty, vagueness, and imprecise information. Fuzzy logic has always been the essence of research-based media decision making. Unlike traditional binary logic, which forces inputs into strict Boolean categories like ‘yes/no’ and ‘0/1’ (or in the case of radio, listener/ non-listener), fuzzy logic enable us to process ambiguous information and make decisions based on flexible rules rather than rigid thresholds.

So if we take these behavioural learnings from BrandMapp 2025 we can develop a holistic weighting factor (Digital Listening Amplification Factor) for the same basket of stations in BRC RAMS. Using DIGILAMP we can get a reasoned estimate of the total unduplicated listenership to these stations today.

The question we need to ask is not so much whether this DIGILAMP weighting technique is right or wrong, but rather “is the fuzzy logic less wrong than using 2-year-old audience data from an obsolete listening diary methodology?”.

The simple reality is that radio listenership is not declining. Multi data-source analysis confirms the evidence of our own ears. Radio listening is increasing but it is morphing into a complex and layered listening behaviour pattern. The hills may no longer be alive with sound of analogue radio, but the audioverse is certainly alive with the sound of curated audio. And nobody can  curate audio better than a radio station.

If you build it, they will come. But if you build it on radio then they will know when to come and where to go!

It’s time to stop hanging on to our radio diaries and say goodbye to Julie Andrews. Or at least get her a few of her favourite things. Like a mobile phone and a set of wearable tracking devices. Then she wont feel so bad.

That’s how you solve a problem like Maria. And that really is sound fuzzy logic.