The radio diary service, whether Arbitron or Nielsen, has been “currency” for over 60 years. Think about that — what can you think of that’s around today, essentially unchanged for over six decades? Sure, the diary’s design has been updated from time to time and the method of recruiting potential respondents went through a major upheaval thanks to the demise of landlines and the near-universal penetration of cell phones. Still, the use of a paper diary to measure seven days of radio listening has survived, which is a tribute to how well it works. But all good things must come to an end someday.
On May 28th, Nielsen held a webinar to talk about potential changes to the diary service, some of which can be considered radical. This week, I’ll set the stage and next week’s column will offer my thoughts on the specifics of Nielsen’s proposed changes.
Consider how we arrived at this point. Long ago and far away, diary service response rates, after adjustments, were in the 30-40% range. Other than government surveys, which have now dropped to that level, that was excellent despite what the Media Rating Council said. Today, response rates are typically in the single digits.
The world has moved online. While there were attempts around 2006-07 that I’ve noted previously in this column, Nielsen has implemented an online version of the diary, named mSurvey, this year. Per Nielsen’s disclosures, about 10% of the sample will be using the mSurvey, at least for now.
A Brief History of Diary Service Competition
For so many years, radio was a powerful medium. Operators made plenty of money and while they would gripe about the price that Arbitron charged, if the dollars rolled in and margins stayed high, they could deal with it. Competitors came and went. Some arrived around 1980 and for those of you in my demo group, names like RAM, Burke, and Trac-7 might trigger a brain cell or two.
Then there was Birch, which was probably the strongest competitor that Arbitron ever faced. I was part of Birch in South Florida when the announcement was made in late 1991 that the company was shutting down. Later, Nielsen took on Arbitron with an ill-fated diary service before eventually buying the company in 2013. Frankly, the diary has a hell of a track record.
But it’s 2026 and times have changed. The webinar spoke about profound methodological changes that will be here very soon. Nielsen can talk all they want about testing, but barring complete disasters, these changes are a foregone conclusion. And while the proposals are grounded in methodological issues, like any company, it also comes down to money. The diary service costs more to run every year while the radio industry continues to contract.
What Nielsen Is Proposing — And What’s at Stake
Today, Cumulus is out while the company challenges Nielsen’s sales policies on antitrust grounds. Sure, the judicial issue is the accusation of “tying” the purchase of local markets to accessing Nationwide at a reasonable price, but Cumulus stated that the ROI on Nielsen data doesn’t work financially in more than half of their markets. Other companies, such as Good Karma, have walked away from Nielsen, but iHeart and Audacy continue to subscribe.
What Nielsen is proposing is both radical and evolutionary at the same time. For non-PPM metros, changing the seven-day paper diary is big. Moving a proportion of the sample to a two-day online questionnaire — 15-20 minutes at one sitting — with modeling to build a seven-day cume is radical. Using sample from different vendors rather than a probability sample is even more radical. However, modeling and non-probability sampling have been around for a long time. Nielsen, and Arbitron before them, didn’t feel comfortable going in that direction and became laggards — but probably in a good way — so what might have been evolutionary over the last two decades or so is now a radical shift.
I say “good way” because the industry prizes consistency over just about any other ratings outcome. Walt Sabo has harped on the industry’s dependence on AQH audiences — small — against cume audiences — large — and he’s right. I’ve said the same thing for years. Regardless, the industry’s currency has been one that is subject to fluctuations, sometimes wild ones, based on a couple of diaries. With radio’s slow but consistent declines in audience, a couple of diaries — or meters, for that matter — can kill a station or even a cluster for months with sometimes dire financial consequences.
There is a bigger issue to consider: the more Nielsen makes the measurement system simpler for respondents, the more they invite competition. If Nielsen says they can get a seven-day cume from two days of listening — similar to the Birch system of 35 years ago — and use convenience samples with acceptable results, anyone can do it. Hire a couple of good statisticians to build the models or train an AI model. And if anyone can do it, a new competitor can charge less than Nielsen, perhaps a lot less. Just roll that around in your mind for a moment.
Let’s meet again next week.
Barrett Media produces daily content on the music, news, and sports media industries. Sign up for our newsletters to stay updated and get the latest information right in your inbox.
