Arbitron PPM President Pierre Bouvard: “It's Radio's Turn To Eat At The Adult's Table” (08/01/09)
By Reed Bunzel, Editor-in-Chief
For many of the 30-plus years that Arbitron has used a diary methodology to track radio listenership, many in the industry have had issues with the accuracy of the reports - and the process. Sample size and response fluctuations have led to ratings wobble, certain demo and ethnic groups were over- and under-represented, reported listening was exaggerated in certain dayparts, and a good deal of cumulative listening was not reported at all. Still, in an imperfect world where measuring out-of-home radio listenership is critical, the diary arguably was, for all its faults, the best thing going (AccuRatings and Birch Ratings notwithstanding).
Enter the Portable People Meter. In the late 1980s and early '90s, radio broadcasters urged Arbitron to look into the practicality of implementing a new, electronic measurement tool that would obviate the need for the diary altogether. The theory was that survey participants would carry a portable device that automatically detected radio signals and transmit that data back to Arbitron's central processing computers. No more listenership estimates, no more pencils and paper, no more issues over whether a station's quarterly ratings reflected actual radio audience. This new device would deliver a clear picture of actual marketplace media consumption, and advertisers - increasingly focused on accountability and return on investment - finally would have data they could trust.
Now, some 15 years later, the radio industry is still waiting for the roll out of the system known as the Portable People Meter. Why such a long delay? Well, for all practical purposes, it's a case of “the devil you know vs. the devil you don't.” Radio executives, managers, programmers, and salespeople have lived with the diary for their entire careers, and consolidation was built on the backs of “sticks” with quantifiable ratings and revenues. Early PPM tests showed some inconsistencies with the diary-fed quarterly reports, and suddenly the industry was faced with a grim reality: What if, all this time, the numbers that were our life's blood were just a little bit skewed?
Since the results of the first U.S. PPM tests were released in 2001, many radio folks have come to grips with the differences that this new methodology might impose. Still, before adopting a new methodology that could significantly alter net revenues and cash flow, the industry collectively asked Arbitron to conduct one more market test, this time in Houston. The ratings giant agreed - with the caveat that, once this test was concluded, it was time for radio to move past the trials and get with the program.
That's where Pierre Bouvard comes in. Named president of Arbitron's PPM division in early January, he has overall responsibility for the introduction of the measurement system in the U.S. Affable and confident, Bouvard is a good choice for the task: He knows the product, he's a long-time Arbitron veteran, and he knows the ratings world inside and out. He's also well versed in the industry's lingering concerns about the “unknown devil,” and clearly took on this new role to bring the hold-outs into the fold.
Prior to his current position, Bouvard was vice president and general manager for Arbitron's Worldwide Media Information Services. In this role, he was responsible for all sales and marketing for domestic radio station services, as well as for all sales and marketing for those same services internationally. Throughout the years, Bouvard has served Arbitron in various sales and management capacities in Chicago, San Francisco, New York, and Dallas; he left Arbitron for a six-year period, during which he was executive vice president of Coleman Research, one of the premier strategic consulting firms for radio stations.
Bouvard is co-author of Radio Advertising's Missing Ingredient: The Optimum Effective Scheduling System, and created Arbitron's series of industry studies covering radio, outdoor, airport advertising, Internet broadcasting, and cinema advertising. A graduate of Northwestern University, Bouvard holds a bachelor's degree in radio, television, and film, and is a graduate of the Advanced Management program at The Wharton School.
Radio Ink saw the start of the Houston test last month as a good opportunity to speak with Bouvard about the logistics, benefits, concerns, and ultimate roll out of the long-awaited PPM.
The Portable People Meter has been tested and retested in several forms in several markets in the past few years. What are you expecting from the Houston test that confirms or refutes the results of earlier trials?
The radio industry, through the Arbitron Advisory Council, asked us to do a test in another major market that would have a large proportion of Hispanics and Spanish speakers. The key question was whether we could recruit a good Hispanic sample and have a good proportion of those that are English-primary Hispanics and Spanish-primary Hispanics. We don't anticipate any surprises, because the electronic radio measurement tests - which we've done in Philly and Wilmington and Manchester and London and Quebec - all show very similar findings. The panel went live July 1; we hit our 2100 sample target on time and on schedule.
How long does a typical panel of participants run?
On average, folks are in the panel six months to a year. That's the whole beauty of a panel: Compared to the diary, the PPM panel has five times the weekly sample. That's why the data is so stable and reliable, and has so much less bounce. Also, every month we freshen the sample by around 5 percent.
Five percent of the panel rolls off and 5 percent rolls on?
Yes, so by the end of the year, 60-80 percent of the panel is fresh.
There's been some criticism that the PPM panel is smaller than the diary panel. Can you address this issue?
Right now, in a typical radio market with 3,000 diaries in the book, we have a sample of 1,000 diaries a month. That breaks down to 250 a week. In the diary, when you start cutting down the time period, your sample size gets reduced dramatically. That's why we have long survey periods - because we're waiting to collect enough diaries for a reliable sample. With the PPM, for that same market, the panel size would be 1300-1400, and it's the same number whether it was yesterday, last week, or last month.
How often will PPM reports be issued?
The new book will be for one month. The major issue advertisers have with radio today is that the audience data takes too long to publish. If I spend $10 million on the Internet on advertising, next week I have a full report on how many people were exposed to my ad. If I spend $10 million on TV or cable, I have a full report on how many people saw that campaign within weeks. In radio, I wait five months, and many advertisers say that's why they don't use radio anymore. Who would wait five months to see the impact of an investment? Imagine if you bought a stock on Wall Street, and the first time you knew how your stock was doing was five months later. Reports will come out a week or 10 days after the end of the month. Right now, if the book for L.A. is released this week, and you're in a medium-sized market, you might have to wait another month until yours comes out. With PPM, all markets come out pretty much on the same day. This is a dramatic improvement in the speed of the data.
Will overnight trends be available?
You will not see it for buying and selling radio, but as a programmer you will most likely use it as an informational tool. Agencies have said to us, “Radio is made up of formats, so the profile and the size of the audience is very consistent week to week, day to day. There is no need for me to buy and sell radio based on overnights, because you don't have the American Idol finale.” Agencies have overnight ratings in local TV markets, but they're not buying based on overnights. Buyers and sellers will be using the monthly data.
Have most radio broadcasters reconciled the fact that their numbers under the diary will be different than their numbers measured by PPM?
We have done so many tests that the smart people in radio know there are no surprises. For a long time, the smart people in radio have known that their cume audience is really twice as big as Arbitron reports. Smart programmers talk about phantom cume, a number that's usually 50 percent more than what people write in the diary. And they were right. PPM shows the average station doubles its cume. PPM also shows higher listening in middays, afternoons, nights, and weekends. No surprise. The big stations are still the big stations and the medium stations are the medium stations. You don't have significant rank jumps.
Still, there has been criticism that PPM will miss certain listening, at certain times of the day - such as morning drive.
Electronic measurement is simply a more accurate reflection of how people really listen to radio stations, rather than how people write stuff down. The fact is, PPM doesn't miss morning drive; it shows that people are overestimating morning listening in the diary. In one study, we called back diary keepers. We held their diary in our hands and said, “I saw that last Wednesday you listened to radio from this time to this time. During this period, how much of the time could you actually hear the radio?” Depending on the person, the answer was 50 to 75 percent of the time - and this was only in morning drive.
So this research suggests that the diary exaggerates some radio listening?
What it's saying is that in the morning, because people are moving around, they're approximating in the diary what they're really listening to. The PPM, however, is picking up how people actually get exposed to radio. Therefore, PPM does not have the exaggeration of morning drive as measured in the diary, but you have higher middays, afternoons, nights, and weekends.
If the cume, average quarter hour, and time spent listening numbers will be different, are radio stations and agencies prepared to price their time accordingly?
We'll definitely see a one-time change in cost per points. In Canada, when our partner BBM implemented PPM, they provided a kind of grid to the agencies to show what PPM was under the old system, and what it would be under the new system. The agencies found this very helpful.
Won't this play up the fact that they've been over- or under-paying for radio for decades?
They're used to it. For 30 years, they've been transitioning TV markets, first to set meters and now to local people meters. This is no big deal for them. The good news is that radio is not the guinea pig on transitioning to electronic measurement.
With such noticeable differences between diary and PPM methodology, why don't advertisers and agencies take greater issue with today's ratings information?
They do. When you take a big step back, the big issue is accountability. Pick up any advertising industry trade publication and all you see are articles on return on investment and accountability. People want to know whether they can trust the credibility of the audience measurement. They will only spend money in media where they think they have credible audience measurement. In 2004, all advertising was up 10 percent; 16 measured media were up while two were flat or down. Those two were network radio and national spot radio. I think everyone recognizes that whether morning drive was up or down is immaterial; radio needs a more secure place in the media plan. We need to improve radio's accountability, and part of that is improving the perceived credibility of the audience measurement.
Can measurement credibility be directly linked to accountability?
Radio needs to get a bigger share of the regional and national ad budgets. If you talk about accountability and ROI, locally radio hits it out of the park. Why? Because the car dealer can actually see the customers come in. The furniture store owner can see the dinette sets going out the door. But, the further you get from the radio station, the lower radio's share of advertising gets. Radio as an industry gets around 25 percent of local ad revenue, but if you start driving out of town, that share drops to 2 percent of national advertisers. That's really where the business proposition is: Does radio want to grow national? Does radio want to get Procter & Gamble money? Does radio want to keep automotive national money?
What degree of cooperation have you received from the radio industry and other media when testing the PPM?
Broadcasters have been talking to us for 20 years about the weaknesses of the diary and the need to move on. In Houston, we have 91 percent of viable local media encoding their signals for this test. We have 16 out of 16 TV stations, 47 cable networks out of 47, and most of the radio broadcast groups. We see a very strong mandate to move forward, not only from radio but also from TV and cable. The radio industry has said, “Let's get on with the show.” [Clear Channel Radio CEO] John Hogan recently urged radio to embrace electronic measurement, saying we must be more accountable to the advertisers. Broadcasters are recognizing that to grow radio's share, we need to improve audience measurement credibility.
With PPM measurement, will station marketers have to rethink how and where internal promotion dollars are allocated?
There definitely will be some paradigm shifts. Radio will still need to market and promote itself. Awareness creates listenership. If a consumer doesn't understand what a station does for a living, or what it stands for, most likely that person won't give it a lot of “behavior.” Just because supermarket scanners electronically measure actual consumption doesn't mean that P&G stopped advertising and promoting. Stations still will have to brand, they will have to stand for something, they will have to be understandable, and they will have to be entertaining.
Where would you expect to see some changes, at least initially?
Some huge benefits will come to radio instantly with PPM, like the speed of data. Going from a 5-month turnaround to a 10-day turnaround will be huge. There's a whole series of accountability and ROI issues that will improve. Radio will be able to say, “We drive retail traffic,” and then prove it. In Houston, right now we're encoding Gap, Best Buy, Gallery Furniture, and Old Navy, so you can see how many of your listeners went to one of those stores over a month or two. That's pure ROI - and that's very exciting for radio.
Will PPM affect the way programming and commercial breaks are scheduled?
If you look at thousands of diaries, you'll see that most people say they start listening to the radio at the top or the bottom of the hour. That's just how people fill out a diary. But how do people really listen to radio stations? They start and stop listening all around the clock. Programmers look at those diaries and say, “If people start listening at the top and the bottom, I'm going to shove all my commercials into quarter-hour two and quarter-hour four.” With PPM, those commercials can be liberated, because the reality is people start and stop listening all over the place. Most likely, you will see more stop sets of shorter duration rather than huge stop sets hidden away in quarter-hour two and four. By creating more stop sets of shorter duration, it's better for listeners, because you can track them minute by minute and see that they hang in there. It's much better for advertisers, because they can be in a stop set of three or four units, as opposed to eight or nine units.
How might almost-instantaneous results change the way contests and promotions are conducted?
The way PPM will affect promotions will blow people away. Right now, when radio people give away concert tickets, they don't think twice about it. If you pull the Maximiser run and look at the week or the day of a contest to see if it had any impact, it's hard to tell. The audience numbers bounce all over the place. But with PPM, you can see the impact of those promotions. In Philadelphia, when WMMR was giving away Stones tickets, you saw huge spikes on those dates. We've never been able to track that before. If I'm the sales manager, and I know that the audience doubles when we're giving away concert tickets, my rates also will double. All in all, it will give the promotion folks much better understanding of what works and what doesn't.
What assurances do you have that respondents will actually carry the PPM with them to work, to dinner, to the beach, to the mall?
The beauty of PPM is that it is the only measurement service in the world where you know, minute-by-minute, whether people are doing what they're supposed to be doing. How do you know someone filled out the diary yesterday? You don't. But the PPM has a motion detector, so we know whether they were carrying the device. The current data shows that the average carry time is 15-16 hours these days, and this is very consistent across key demo and ethnic groups. We know that Hispancs, blacks, and others comply at about the same level. Another advantage of PPM is that non-compliance doesn't depress the ratings. If someone doesn't fill out a diary page, there's a blank page that doesn't contribute a lot even though we do ascribe some data to the ratings. If someone didn't carry the PPM, that day we don't count them. It isn't zero listening; it's just that there's no participation that day.
Do you think people will wear a pager-like device when they're already carrying cell phones and iPods?
When we're getting 15-16 hours a day of compliance, we don't see any issues with the look and feel of the device. Some people have said we should make it look like a cell phone, or make it part of a cell phone. The fact is that people don't have their cell phones strapped to them when they're in bed at night. It's easy to say “make it like an iPod,” but right now, we are getting spectacular participation rates, by all age and race groups.
Still, response rates for all market research are declining. How is Arbitron dealing with this problem?
In the PPM world, the term for response rate is sample performance index: SPI. In Houston, the household SPI - meaning whether we got data back from someone in the participating household - is around 46 percent, while the persons' SPI is around 29 percent. We're very happy with those numbers. On the other hand, the hard part in trying to maintain response rate in a one-week diary methodology is the week comes and goes very quickly. You're in or you're out, you filled it in or you didn't, and our impact on that person is over. If I have you in a panel, I can work you. I can call you, I can give you points, I can whisper in your ear. You're in my panel, so we have more of an opportunity to impact response rate and try things than we do in a quick, in-and-out diary methodology.
Close to 10 percent of all U.S. households are cell phone-only. Are you having difficulty reaching these people?
The great thing about our current execution is that all households are participating, and the latest number I saw was 12-13 percent of the panel is a no-land-line-phone household.
What steps have you taken to evolve the technology from when PPM first was introduced?
We've been modifying this every year. Believe it or not, in Houston we are on generation 10. The PPM of 10 years ago has been significantly enhanced and modified 10 times. Still, there's a myth in radio that PPM is old technology. That's nonsense. I can show you our patent room where you literally can see, year-by-year, all the new functionality and new technology that has gone into PPM - and we don't intend to stop.
How does PPM work with HD Radio - both the main channel and secondary multicast channels?
We've tested it, and it works spectacularly. U.S. radio is expanding its distribution; no longer is radio about tall towers in big fields. It's about putting content on HD Radio, on a cell phone, and on the Internet. The only way you're ever going to measure your streaming audience compared to HD, compared to cell phone, is with PPM, because we can put in one encoder for each of those, so you can get very accurate numbers on what percent of your audience is coming from your different distribution platforms. The beauty is, you're encoding your different streams with a different code, and the PPM can easily pick it up.
Recently there's been discussion that PPM will expand radio's scope from just a frequency medium to now include reach. Can that be done without sacrificing results?
Erwin Ephron is the father of modern media planning and one of the most respected agency-advertiser gurus on Madison Avenue. Erwin's position is that until now radio has done itself a great disservice by positioning itself just as a frequency medium.Under PPM, the average station's reach doubles, which means it will end up in far more media plans. Buyers can more easily justify a larger share of budget for radio, because they can accomplish significant reach and frequency goals vs. TV at a fraction of the cost.
Since broadcasters already have price issues with the diary, are they reluctant to pay more for PPM?
We will give people the best possible deal for electronic measurement. Twenty years ago, when we switched our TV clients from the diary to meters, their rates tripled. We do not anticipate anything near that for radio, because we have been working very hard to make this absolutely the best possible deal vs. any other media that has gone from paper and pencil to electronic measurement.
Assuming the results of the Houston test are positive, do you have a schedule in place to roll out PPM?
We will commercialize PPM as fast as the radio industry wants it. We could conceptually commercialize Houston by mid-year next year, if the industry wants to go that fast. Everything we do is market-driven. We are meeting with our large customers in major markets and asking them what they would like to see. We're hearing that broadcasters want us to start in the biggest markets. But, we will fine-tune roll-out schedules in direct consultation with customers.
Will you overlap the diary with PPM measurement at first, or do you prefer to shut down the diary in those markets where PPM has begun?
We'd want to get input from customers on that. We've spoken with agencies and station folks on the TV side who have gone through the same transition, and we get the strong sense that people want to get on to the new thing. Initially, people may run a diary or two in parallel, but based on the experience of other media, I think they'll want to switch over without a lot of overlap.
With all the testing, feedback, retesting, and anticipation, what have you learned about PPM - and what would you tell broadcasters who still have concerns about it?
I'll repeat what we hear from radio's largest advertisers: “It's radio's turn to eat at the adult's table.” Radio deserves a larger share of the national media buy, and we want to give them the tools to get it. We think the next few years will be very exciting, because as we transition to PPM, buyers and sellers of radio will work together on a bunch of issues. Nothing we say will convince the skeptic. Radio's largest customers must sitt down with broadcasters and say, “Let's talk about increasing radio's share” and “Let's talk about the tools we need to do that.”
Comment on this story
|| From the Publisher