The ratings game

I was a post-teen Arbitron diarist -- and all I got was a lousy 4 bucks.

Topics:

The ratings game

SHE was dead.

Late one Saturday night in the spring of 1999 I was editing the latest issue of my college paper. As usual, the radio was set on Orlando’s WSHE 100.3 FM. I was fanatical in my devotion to the station, primarily because I’d never before found one that so perfectly matched my taste. Plus, it had cool DJs and often sponsored great concerts and events. Sure, they overplayed songs and had too many ads, but what station doesn’t?

Around 11:30 that night, REM’s “It’s the End of the World as We Know It” came on. I didn’t catch the hint. Next up was a song I’d never heard before on that station. Then another. Then the death knell: a station ID, identifying it as “the new Cool 100.”

I fiddled with the dial. Maybe the signal was gone and we were picking up another station. Or maybe they were promoting a new station temporarily on SHE’s band. Something. But hours later, there was no change. I called the station. The person who answered the phone confirmed my fear: They’d switched formats. Gone were the Barenaked Ladies and Sarah McLachlan; in were groups I couldn’t even name if I tried. No notice. Nothing. Everything I’d grown to love, DJs and all, was gone. SHE was dead.

SHE may have been just a radio station, that didn’t mean WSHE’s demise was any less of a surprise or any less wrenching for me. The station’s Web site soon offered an explanation: “We are sorry that you are disappointed with the format change on WSHE. The low ratings on SHE forced us to make a change to a format that has a larger listener base.” The Orlando Business Journal gave the details: SHE was 14th in the market, pulling in $3.5 million a year. Another station had recently changed formats. Its previous, now-abandoned oldies format had kept it in the Top 10, and it had twice SHE’s revenue. So SHE’s parent company, Clear Channel Communications, which owns 900 radio stations in the country, leapt at the opportunity to switch to oldies and grab some of those now-floating listeners — and the ad dollars they brought in.



Of the 22 markets that Radio-Info.com tracks, the site lists at least 10 changes in format since September. Elsewhere, Radio & Records Online reported on Nov. 20 that three different stations had changed formats the previous Thursday and Friday alone. Radio stations come and go now; this is a nationwide reality. And the changes are all over the board: jazz to modern rock, news to all ’80s, oldies to alternative. But one thing’s for sure — their reasons all sound familiar. Radio-Info provides detailed descriptions about the format changes they list (and they definitely miss some, including WSHE’s), and low ratings and holes in the market are frequently cited as rationale for the switch.

SHE itself was only 2 years old when it went under, and it had replaced Orlando institution WDIZ. Still, I was distraught and wrote about how pissed off I was. Later, grief took the form of reaching for that familiar preset button and finding nothing familiar. Finally, I reached the stage of acceptance and rationalization: Hey, at least I would hear more pop music.

But even now, a year and a half later, I’m still annoyed. My radio station was killed because its ratings were low. Yet everyone I knew listened to WSHE. It seemed like its events were packed and the advertisers fat. To me, that meant that whoever compiled those ratings had asked the wrong people. It seemed like the station format was killed because a handful of random people liked to listen to oldies and other formats instead of alternative/adult contemporary music. And that really irritated me.

About six months ago, just over a year after the death of SHE, I received a random telephone call. It was Arbitron, the company responsible for “measuring radio audiences in local markets across the United States,” according to its Web site. I’d been randomly selected to record my listening habits for a week, which would help to form the spring ratings. Was I interested in participating? Hell, yeah. It was payback time.

The process was explained, and then the person on the phone expressed concern that I, a 22-year-old male, wouldn’t return the diary. “We can count on you to put it in the mailbox?” the representative asked repeatedly. “Yes, yes,” I said, exasperated.

Apparently, “men in [my] age group” aren’t well known around Arbitron for returning diaries. Mine, I assured them, would be in the mail. I was told I’d be paid a small amount for my participation, but I wasn’t supposed to go out and buy a car — just a cup of coffee and a newspaper.

Within a few days, I received a letter. “Thanks again for agreeing to take part …,” it said. The letter rehashed what the rep had told me and reiterated why my participation was so important. A brochure explained the process and showed a group of people in 1980s clothing sitting by radios and dutifully filling out their diaries. The most fascinating part of the mailer, however, was a $1 bill stuffed into the envelope. A slightly wrinkled, I-just-pulled-this-out-of-my-wallet bill. My compensation? They had to be kidding.

A few days after that, I received my diary and another letter. Also enclosed, according to that letter, was another “token of our appreciation” — three more dollars. A little crisper, but still three loose $1 bills. I was beginning to feel awfully cheap.

I looked at the diary, a small paper book. It was preprinted for seven days; a little chart showed me how to fill out the daily grids. Every time I heard a radio, I was to record the start and stop time, the station (“call letters, dial setting, or station name”) and the place where I was listening to the station. Interestingly, “listening” was defined as “any time you can hear a radio — whether you choose the station or not.”

On the first day, I woke up around 7 a.m. and promptly forgot to turn on the radio. Once I remembered that I was supposed to be recording my listening habits for the ratings, however, I immediately turned it on, debating over what morning show to listen to. I went with the one I listened to most often. I recorded the time. Later, I took my diary to work, where I dutifully recorded my listening there. This was easy.

But I was having a hard time figuring out how I was going to extract revenge. Fabricating my listening habits never even occurred to me; instead, I realized this was my chance to support the kind of stations I liked by listening all day, every day. I decided to extend that to stations outside my area. I live in Chicago now, but after I moved here, I continued listening to an Orlando station almost every day at work via the Internet. Gradually, I stopped. But now, diary in hand, I was going to listen again — I wanted to help boost its ratings so it wouldn’t die an agonizing death like SHE. Would listening to a station via the Internet count? I wrote it down anyway, giving a lot of information in the tiny box: “104.1 WTKS Orlando via Internet.”

During the rest of the week, I tried to be good. Someone gave me a ride home one day, and being a good diary keeper, I asked what station we were listening to, and recorded it. But my plan to listen all of the time never materialized. My listening was definitely the heaviest on days one and seven — the days I was most conscious of my participation. Sunday, Monday and Tuesday, my listening was rather gratuitous — “Oh, I should listen to the radio now.” I’d flip it on, record the hour or two. All in all, it was roughly 30 hours of radio time over seven days.

Before I mailed the postage-paid diary back to Arbitron, I had to supply some demographic information: my age, sex, location, employment status, education, number of children in the household and income. I also had to tell them what station I listened to the most six months ago. When I first received the survey, the big blank area on the demo page that said “Your opinion counts” called out for a screed against Arbitron for leading to the death of my beloved SHE 100.3 FM in Orlando. But I left it blank. I was tired of the damn diary. I photocopied it and dropped it in the mail.

And then I began to think about the whole process. Specifically, rumors of Nielsen families came to mind. Much like Arbitron, Nielsen surveys families on their TV viewing habits. I’d heard that, compared to those families whose viewing is recorded by an electronic device, those who self-report their viewing habits generally exaggerate about what they watch. (They don’t admit to spending a lot of time on fluff, instead sending in journals that claim they’ve watched only the news and insane amounts of PBS.) Was my self-reporting really accurate? How exactly can Arbitron survey a few random people with such seemingly inaccurate methodology and use that to issue what amounts to death warrants for low-rated radio stations?

Arbitron, as far as I knew, relies solely on people like me to report accurately the stations we listen to. And as a result, stations — and more significantly, their parent companies — make decisions that affect all listeners. I e-mailed Thom Mocarsky, vice president of communications at Arbitron, to find out exactly what Arbitron does and why they do it that way. As it turns out, it all comes down to math.

Like all of the Arbitron diarists, I was selected at random. Actually, my unlisted telephone number was selected at random. And that random sample is, Mocarsky said, “representative of the local market that sample is drawn from.” But is it? How do you just pick random people and know they’re like the whole population?

For one, reports are adjusted to make sure demographics participate equally. “When we tabulate the data, we ‘weight’ any returned sample segment that is slightly under or over representative so that it acts as if it were perfectly proportional.” He offered an example: “If we need 340 diaries for men 18-24 and we only got 322, we make those 322 diaries have the mathematical weight of 340 diaries when we compute the ratings.” Mocarsky said that those percentages — X percentage of women, Y percentage of people 18 to 24 — is reported with each survey.

But more significantly than that, I thought, how can you just pick a few people and claim they represent everyone? I sent an e-mail message to Dr. John Rasp, a statistics professor at my undergrad alma mater, to see if he could help explain. First, I asked him about the methodology of picking a “representative” group. Finding “a small group that gives a cross-section of the … whole,” he said, “… can never be done perfectly — but can be done close enough for practical purposes.” And Rasp, an associate professor of decision and information sciences at Stetson University, says Arbitron’s method of random selection of telephone owners is “a fairly close approximation” to the population.

Also significant is who participates. Mocarsky says that, of those selected to participate, “38.7 percent of those people sent back a diary that was complete enough for us to tabulate in the ratings report” — which is “a very good response rate for our type of survey methodology — a seven-day, self-administered diary.”

Fair enough. But even though it seems logical, it still rubs me the wrong way. What if the wrong group was selected? Don’t you get a different group of radio listeners every time? Rasp drew an analogy: If I were in charge of taste-testing a 700-gallon vat of a beverage, I wouldn’t have to drink the whole thing, or even 70 gallons, to know it was OK. Rasp says, “A mouthful or two will do. Same principle applies to statistical sampling. Assuming the ‘vat’ of people you’re sampling from is well mixed (and that’s the purpose of the randomness), you just need a ‘mouthful’ or two — a few hundred, maybe a couple of thousand, people.”

I turned the metaphor around on him and said that people aren’t like a homogeneous liquid; they’re more like a barrel of mixed nuts. He agreed, but again pointed to the numbers. As long as they’re selecting enough people, it’s OK. “[I]t’s the size of the scoop that matters, not whether the barrel has five or 50 tons of nuts.” Mocarsky told me that of Chicago’s 7,147,300 people over the age of 12, they had 6,095 diaries — one for every 1,173 people. Which, I confess, doesn’t seem too bad at all; I’m OK with representing 1,173 people. Plus, variation in two different samples is accounted for and quantified by, Rasp says, margins of error, which Arbitron says are different for each survey but are offered with each report.

For Rasp, the “key concerns” lie not with the margin of error, but with “honest reporting of data.” And he’s not buying my accuracy, or anyone else’s, in recording our own listening habits. “Folk[s] do not recall their listening habits accurately. Not everyone is willing to keep diaries of their listening. And those who do are not necessarily like those who don’t.” The results, Rasp says, “are only as good as the input.”

Mocarsky defends their methodology and experience. “We invest a lot of money ‘researching the research,’ constantly improving our methods, and working hard to get a telemarketing-weary American public to take part in our surveys. We constantly work with our customers to address their interests and concerns.”

And they disclaim the heck out of themselves. “[L]est you think we’re pulling the wool over anyone’s eyes we cite 16 specific categories of limitations — including this one ‘Diaries, or portions thereof, may be completed improperly if the diary instructions are not followed by the diary keepers,’” Mocarsky says. “Within the stated limitations of our methodology, we do a good job of tracking local market radio audiences in the United States.”

Plus, they’re audited and “accredited by the Media Ratings Council, an independent, industry-supported organization that is charged with seeing that the audience estimates used by the broadcast and advertising industry are ‘valid, reliable and credible.’”

But what about the cash, essentially a bribe to get me to fill the diary out? Predictably, Mocarsky says Arbitron includes money to help encourage participants to return their surveys. Even the cash gifts are calculated moves: Because certain demographic groups — like my own, 18- to 24-year-old men — “do not respond as readily” as others, Mocarsky says, Arbitron gives us more cash. “We regularly calculate the cost of the cash incentives we enclose against the benefits of the increased response rates,” he says. And Rasp says that such rewards are a “fairly common practice” that “does serve to increase the response rate moderately”; he doesn’t think it affects the response negatively.

So all this is making sense. But what about my beloved SHE 100.3 FM? Why did Arbitron sign its death warrant when so many people I knew were listening?

Apparently, people my age were listening, but no one else was. “In the fall of ’98 — the station had a total audience of 148,800 people — or 13.5 percent of everyone 12 [and up] in the market,” Mocarsky said. Among 18- to 34-year-olds, SHE was No. 6, and No. 4 for women 18 to 34. But for adults 25 to 54, it wasn’t even in the top 10. And the 18- to 24-year-olds I insist were listening? They’re only 11.5 percent of the population, Mocarsky says, and this past spring, 11.9 percent of the surveys came from that group. “[Y]our peers have had appropriate representation in our surveys.” Clear Channel obviously looked at those numbers, and then those of the station that changed formats, and decided to make a switch. A sound business decision, albeit one that affected me directly.

So, OK, maybe no one was listening. But not this time. How did my patronage of the Orlando station via its streaming audio on the Web affect its ratings? “The issue of how Internet listening is reported by Arbitron is a subject of much debate among our customers,” Mocarsky admits. “If you lived in Chicago, and you entered enough information for us to identify that Orlando station and to know that you listened over the Internet, the station would get credit in the database we use to tabulate the Chicago Local Market Report.” In other words, my ratings didn’t do anything to help the station in Orlando — even though I was listening to their local and national ads. But if enough people in Chicago had listened to that station via the Internet, it would have appeared right alongside traditional stations in Chicago.

That hasn’t happened yet, though: “[T]o date, no station has gotten enough Internet-based entries to meet the minimum reporting standards for any of our services,” Mocarsky says. But my listening did count for something; those hours were included in the “total radio listening” report for Chicago.

Well, I tried, making the best use I could of my chance to affect radio history. And there’s always next time. The first letter I received from Arbitron told me I probably won’t be asked to participate again until 2042 — when I’ll no doubt be listening to oldies like Eminem and ‘NSync. Then it’ll be time for true revenge.

Andy Dehnart is a writer living in Chicago.

More Related Stories

Featured Slide Shows

  • Share on Twitter
  • Share on Facebook
  • 1 of 11
  • Close
  • Fullscreen
  • Thumbnails
    Burger King Japan

    2014's fast food atrocities

    Burger King's black cheeseburger: Made with squid ink and bamboo charcoal, arguably a symbol of meat's destructive effect on the planet. Only available in Japan.

    Elite Daily/Twitter

    2014's fast food atrocities

    McDonald's Black Burger: Because the laws of competition say that once Burger King introduces a black cheeseburger, it's only a matter of time before McDonald's follows suit. You still don't have to eat it.

    Domino's

    2014's fast food atrocities

    Domino's Specialty Chicken: It's like regular pizza, except instead of a crust, there's fried chicken. The company's marketing officer calls it "one of the most creative, innovative menu items we have ever had” -- brain power put to good use.

    Arby's/Facebook

    2014's fast food atrocities

    Arby's Meat Mountain: The viral off-menu product containing eight different types of meat that, on second read, was probably engineered by Arby's all along. Horrific, regardless.

    KFC

    2014's fast food atrocities

    KFC'S ZINGER DOUBLE DOWN KING: A sandwich made by adding a burger patty to the infamous chicken-instead-of-buns creation can only be described using all caps. NO BUN ALL MEAT. Only available in South Korea.

    Taco Bell

    2014's fast food atrocities

    Taco Bell's Waffle Taco: It took two years for Taco Bell to develop this waffle folded in the shape of a taco, the stand-out star of its new breakfast menu.

    Michele Parente/Twitter

    2014's fast food atrocities

    Krispy Kreme Triple Cheeseburger: Only attendees at the San Diego County Fair were given the opportunity to taste the official version of this donut-hamburger-heart attack combo. The rest of America has reasonable odds of not dropping dead tomorrow.

    Taco Bell

    2014's fast food atrocities

    Taco Bell's Quesarito: A burrito wrapped in a quesadilla inside an enigma. Quarantined to one store in Oklahoma City.

    Pizzagamechangers.com

    2014's fast food atrocities

    Boston Pizza's Pizza Cake: The people's choice winner of a Canadian pizza chain's contest whose real aim, we'd imagine, is to prove that there's no such thing as "too far." Currently in development.

    7-Eleven

    2014's fast food atrocities

    7-Eleven's Doritos Loaded: "For something decadent and artificial by design," wrote one impassioned reviewer, "it only tasted of the latter."

  • Recent Slide Shows

Comments

0 Comments

Comment Preview

Your name will appear as username ( settings | log out )

You may use these HTML tags and attributes: <a href=""> <b> <em> <strong> <i> <blockquote>