There was an interesting debate going on in the “Tom Taylor NOW” newsletter late last week over the recent release of a Borrell Associates white paper called Future of Legacy Media.
Borrell’s report argues that consumers are trading time-spent-listening (TSL) of legacy media like newspapers, radio, and TV for digital media brands, and supports that point with the observation that AM/FM radio’s time-spent-listening has declined 30% between 2008 and 2013. (Since radio’s cume has been holding essentially steady, that means radio’s AQH has also declined by about 30%.)
On Friday, Tom Taylor’s readers were asking if that apparent TSL decline was illusory — i.e., merely the result of a research methodology change, Arbitron’s switch from paper diaries to PPMs (Portable People Meters) in large markets. (“The time period coincides, exactly, with Arbitron’s phased rollout of electronic measurement,” one reader wrote.)
As longtime RAIN readers may recall, I have some experience in this area: For almost 20 years (off and on during the period 1980-2013), I ran a market research firm called Strategic Media Research that was for a while the leading music research firm in America. And when the ratings firm Birch Ratings went out of business in the early 1999s, we launched a telephone interview-based ratings product called AccuRatings that had a successful several-year run that is still remembered fondly by some old radio hands.
So, some thoughts and observations:
- Radio’s actual problem with ratings in the 1990s was one of inadequate sample size: With Arbitron using only a few thousand diaries in a given market per book, a 4.0-share radio station could gain or lose up to a full share point per book simply due to margin of error. (I used to have a slide show — literally 35mm slides; this was before PowerPoint — using a “Marbles in a Swimming Pool” analogy that illustrated the effects of standard error on a research finding.)
- AccuRatings was a product designed to solve the sample size problem: Developed with researchers like Amy Vokes (now at Radio One), Gregg Peterson (now at Market Strategies International), and Tripp Eldredge (now at DMR), it was based on a brief telephone interview that allowed us to use a sample size that was typically three to four times larger than Arbitron’s. (And it had a higher response rate than diaries did, so its findings were likely more representative of the general population.)
- Consolidation largely killed AccuRatings: We were successfully measuring most the top 70 markets in America, but as Clear Channel, under Randy Michaels, kept acquiring more and more of our station clients, Randy insisted that those stations drop Strategic Media Research and switch to a research company that he himself personally owned. (We admittedly made a few mistakes in management hiring, too.)
- Arbitron’s introduction of Portable People Meters (PPMs) solved the wrong problem: Diaries weren’t being filled out precisely to the minute (in fact, lots of respondents filled out their diaries in one sitting at the end of the week), and PPMs “fixed” this problem, but because meters were so much more expensive than paper diaries (hundreds of dollars vs. maybe a dime), Arbitron used far fewer of them per market. So the sample size issue got *worse,* not better!
- More importantly, PPMs introduced a much bigger problem: While people’s entries of their listening behavior in diaries no doubt were somewhat inaccurate, they were inaccurate *in radio’s favor*! In the markets where PPMs were intro used, overall radio listening per Arbitron dropped something like 25%. Nationally, using a mix of PPM and diary markets, radio listening appeared to drop 15%.
- I’ve always thought that radio is to be totally commended for maintaining its revenues and spot rates during the switch to PPM data. After all, if Arbitron ratings showed a high decline in AQH yet stations successfully held their rates (by arguing that their audience size hadn’t changed, only the methodology had), that means they achieve a significant *increase* in cost-per-point and cost-per-thousand. “Good job, radio industry, on this one!”, I’ve always felt.
- But, wait: Actually, I see BIA/Kelsey data that says that radio station revenues declined from $17.9 billion in 2007 (when Arbitron was all-diaries) to $16.5 billion in 2008 (the year major markets were being switched over to PPMs) to $13.3 billion in 2009 (the first year post-transition to PPMs). That’s a 25% decline — a decline of $4.6 billion annually — that radio hasn’t fully recovered from yet. As I think about history, I attribute this decline to the recession … but is that an incorrect interpretation?
Total U.S. ad spending decline 13% in 2009, but radio ad spending revenues declined 20% that year. Was the extra 7% decline (i.e., $1.15 billion per year), to some extent, the annual (and perhaps continuing ) cost to the radio industry of switching from diaries to PPMs?
In any case, PPMs were introduced in major markets in early 2008, so Tom Taylor’s readers have the date wrong. By 2008, PPMs were already in place in most major markets, and that’s the year that radio’s TSL (hours per week per person) dipped from 17.6 to 15.2 hours.
As for the 30% decline in radio listening TSL that the Borrell study references, I believe that occurred *after* 2008.
Which leads to a big question: *Why* has AM/FM radio lost 30% of its TSL and AQH since 2008? I don’t think this question has been vividly addressed in any major national research study.
However, it seems to me that it’s due to a combination of factors, the first three all being effects of consolidation: (A) The eventual result of over 15 years of shrinking (or no) marketing budgets for stations. (Imagine if Coke and Pepsi and RC all quit advertising for 15 years to save on marketing costs. Cola consumption would probably eventually decline, right?) (B) The result of over 15 years of increased spot loads. (C) The result of over 15 years of cutting back on talented programmers and air personalities to reduce costs. (D) The growth of satellite radio (now with over 25 million subscribers, representing almost 20% of U.S. households), and almost certainly having a negative impact on AM/FM listening in cars. (E) The growth of Internet radio (now used weekly by over a third of the U.S. population), almost certainly having a negative impact on AM/FM listening in the workplace.
But blaming it on the introduction of the PPM? Nope. Primarily, that was the previous 15% decline (2006-2008), *not* the subsequent 30% (2008-2013).