Arbitron Client Briefing Explodes '5 PPM Myths'
March 25, 2011 at 4:21 AM (PT)
The ARBITRON FEBRUARY PPM Client Briefing took an interesting turn after company executives cited benchmark-worthy performance metrics, updates on its in-person recruitment and ARBITRON's transition to address-based sample. Two executives from COLEMAN INSIGHTS took the virtual podium to debunk five "PPM myths" that are nonetheless being taken for gospel by more than a few programmers.
Chairman/CEO JON COLEMAN cited the first two myths -- that PPM numbers are "much more stable and reliable" than diary data and that you can immediately discern what works and what doesn't with PPM. "While the numbers can be stable, there are still wobbles," he said.
It would be wrong to play a very popular song if it didn't fit the station's brand. 'You have to weigh long-term brand decline against a short-term ratings gain'
Regarding the second myth, COLEMAN noted that the sample size is still too small to make a judgment on a day-after spike or drop. "An individual song or content piece may show no in-the-moment impact, yet it still makes an impression that can lead to more/less listening in the future." The data should only be considered reliable in the aggregate, over many weeks.
What's more PPM measures tune-out far better than tune-in, which means ratings won't spike up every time you do something well. What does register with the PPM. The more significant the event, the bigger impact on ratings. A dramatic event would be a format change, a moderate event would be a franchise contest or a formatic change, a minor event would be a non-franchise contest or minor music change, and a negligible would be a new song or a single news event.
Do Brands Matter?
VP JOHN BOYNE tackled Myth #3 -- that brands don't matter in PPM; it's all about product execution, when in reality you need both strong content and a strong brand; and Myth #4 -- that you have to eliminate everything that causes the minute-by-minute numbers to drop ... and you must increase the frequency of things that cause minute-by-minute numbers to go up.
Despite PPM data that habitually shows listening levels dropping during between-song talk segments and long transitions, BOYNE asserted that the need to maintain high numbers and disguise transitions should be shared with the need to maintain a strong brand.
Citing a study undertaken by STARBUCKS as an analogy -- where they found that long lines during rush hour actually improved the company's mellow ambiance, BOYNE asserted that it would be wrong to play a very popular song if it didn't fit the station's brand. "You have to weigh long-term brand decline against a short-term ratings gain," he said.
In fact, while there will always be listeners who will tune out of talk and transitions segments, they key is to make the talk and transition segments as entertaining and brand-affirming as possible to grow the base of those who do listen to all the elements.
Is Marketing Necessary?
Myth #5 is that since PPM measures actual behavior instead of recall, you don’t need to market as much. Disproving that notion was an earlier study cited by COLEMAN, which found that 1) most listening is habituated; 2) listener rely on pre-sets; 3) there's little scanning and station surfing outside of pre-sets; and that 4) even though PPM does pick up incidental listening, by far the most listening comes from intentional use.
To generate intentional listening, you have to be well known, own a position and build a brand -- which is where marketing comes in big-time.
* "Expect wobbles and uncertainty" in the PPM
* "Aggregate, aggregate, aggregate" the data over weeks before taking action.
* "Don’t confuse correlation with cause and effect; understand what’s truly causing your ratings to go up or down."
* "Assess the value of your content using both ratings and brand value measures."
* "Aim to be a station that has both a strong brand and great content."
Read ARBITRON's entire Client Briefing here.