Sorry, wrong number

Troubled tallies take toll on advertiser confidence

By MICHAEL LEARMONTH

He's been fined by the FCC and dumped by Clear Channel, but Howard Stern's summer ratings are up, way up.

How far up? His employer, Infinity Broadcasting, doesn't really know.

"I saw the numbers in the newspaper," says Infinity prexy and chief operating officer Joel Hollander. "So we have an idea he's doing fairly well."

Fairly well.

For media buyers interested in advertising on one of Infinity's 85 stations, "fairly well," or any number of other adjectives, will have to do. In a move akin to a jeweler throwing out his scale, or a stockbroker trading without a share price, Infinity decided to abandon its contract with Arbitron, the company that provides the primary ratings used to buy and sell radio air time.

The Infinity decision stunned advertisers and even Arbitron itself, but flareups of this kind are becoming more commonplace as tensions mount between media and the ratings agencies that measure them.

For decades, advertisers have relied on "gold standard" currencies provided by ratings agencies such as Arbitron, Nielsen and the Audit Bureau of Circulations. But as viewer, listener and reader attention fractures among thousands of outlets -- TV, radio, print and online -- advertisers are demanding much more precise data, a challenge to the ratings agencies and a threat to the media pecking order.

The result has been some of the bloodiest conflicts over ratings data in decades.

"You haven't seen an increase in sophistication of audience measurement go hand-in-hand with the increasing complexity of media," says Peter Gardner, chief media officer at ad agency Deutsch.

Ratings agencies are tradition-bound and loath to alienate the media they measure -- and that pay their bills. But hundreds of niche television channels and new technologies are carving up audiences, requiring much more precise measurement of smaller audiences across all types of media and turning every ratings point into a battleground.

Advertisers, expected to spend north of $160 billion on television, magazines, radio, newspapers and every other kind of media in 2004, are understandably demanding more accuracy and accountability before they commit their dollars.

For instance, ad agencies are pursuing a number of initiatives that would allow them to measure the actual viewership of a TV spot, not just rely on the overall average rating for a program. In an era of TiVo and zapping, that could mean a significant change in numbers (and in ad fees).

"There's an increased demand for better information," says advertising analyst Jack Myers. "That is spurring discontent."

Virtually every medium has been forced to do some soul-searching this year as ratings flaps erode the credibility of the fundamental currency with which media is bought and sold.

• Fox and Univision have waged a public fight against Nielsen over the rollout of its "Local People Meter" technology in top markets.

• Infinity dumped Arbitron in June after what it calls a dispute over the cost of ratings.

• Tribune Co., Hollinger Intl. and Belo Corp. were stung by newspaper circulation scandals, including some that persisted for years despite repeated audits by the Audit Bureau of Circulations.

• Magazine publisher Gruner+Jahr admitted last year it overstating circulation at several magazines, including YM, Fast Company and the now-defunct Rosie.

• Cable operators, with data from 65 million set-top boxes, are weighing a parallel ratings service that could deliver a serious blow to Nielsen.

Infinity and beyond

On the surface, Infinity-Arbitron was about price; some observers wrote it off as Infinity's latest negotiating tactic.

But Infinity's Hollander says his dispute with Arbitron is bigger than the size of the fees it charges. "We're trying to change a paradigm," he explains.

Radio is still measured primarily with written diaries, a system that is laughably inaccurate, but the industry clings to it because no one wants to be on the losing end of whatever new precision could be achieved through a better system.

No one wants to pay for a new system, either.

A typical written radio diary says people listen to about 20 hours of radio a week. They generally stop and start listening on the hour, and listen to about two stations a day.

But in the real world, people listen to far more stations, more like six during the day, in smaller increments. They start and stop on fractions of the hour and spend far more time, perhaps 40 hours, listening each week.

How do we know this? Arbitron is experimenting with an electronic monitoring system, similar to that used by Nielsen to measure national TV audiences. Arbitron first tried the system in Philadelphia and is planning a yearlong trial of the service next year in Houston.

Resistance to the new system has already begun, threatening Arbitron's ability to make the shift. Infinity, the second-biggest radio chain in the country, dumped Arbitron in June, citing the escalating cost of the service. Entercom, the fourth-largest radio chain, briefly let its Arbitron membership lapse this summer as it renegotiated its contract.

"What's the long-term issue for the industry? Is it People Meters, diaries, telephone retrievals?" Hollander asks. "The existing service has not been able to give us cost certainty on what it's going to be."

Talks between Infinity and Arbitron continue, but in the meantime the radio group contracted with Houston-based market researcher the Media Audit to provide market research. "(It's) just one of the methodologies we will utilize to provide our stations with the highest-quality research needed to compete in their markets," Hollander says.

Don't discount controversy

Electronic People Meters are at the center of the ratings dispute wracking television, pitting broadcast outfits with large Hispanic and minority audiences against Nielsen.

Broadcast knows it must change to compete with the viewership challenge posed by cable, and no one says they don't want the more precise data Local People Meters, or LPMs, can provide.

But a more precise measurement tool requires a more precise sample. Univision and Fox say flaws in the way Nielsen recruits its samples end up undercounting blacks and Hispanics, key demographic groups advertisers want to reach.

Univision, which tried to stop LPMs in Los Angeles in court, has accused Nielsen of not making the effort to get enough large, Spanish-speaking households in the sample, causing a 39% dropoff in reported viewers between the ages of 18 and 34.

"Nielsen has got to fix the sample, it's that simple," says Ceril Shagrin, head of research for Univision and a former Nielsen staffer. "Put a good tool with a bad sample -- you have bad data."

But a district court judge threw out all of Univision's claims and Nielsen says the sample it has recruited is far more comprehensive than the one it replaces.

The proof, Nielsen says, is in the ratings. "We see ratings to popular channels drop and ratings to niche channels rise," says spokesman Jack Loftus. "The People Meter is showing audience trends more diverse than the existing diaries are capable of reporting. It's five minutes here, 12 minutes there across dozens of channels."

Latinos aren't being undercounted, Nielsen says, they've just moved to smaller cable outlets that better serve their needs.

While Univision has pursued its claim in court, Fox has waged a public relations battle through organizations like Don't Count Us Out that could prove even more damaging to Nielsen. Fox, which owns several key UPN affiliates, says undercounting of minorities at UPN will lead to less minority-oriented programming.

Some industry observers fear that if Fox succeeds in introducing even the smallest negative opinion of People Meters in the marketplace, the system, and Nielsen's ability to administer it, could be doomed.

Congress held hearings on the dispute last month and Fox found itself in the bizarre position of asking the government to step in and regulate the TV ratings business. Both houses rejected the plea.

Mediapost editor Joe Mandese expects strong-arm tactics to become more commonplace.

"Little shifts in the method can have a huge impact in terms of market share and revenues and profits," he said. "It's understandable that the management of companies will put pressure on the research to skew the thing more in their favor."

Circulation's pulse

The circulation scandals plaguing newspaper and magazine publishing are a bit different, borne out of publisher pressure to maintain an unrealistic picture of vitality as circulation falls across the industry.

But the scandals also have cast negative attention on the not-for-profit cooperative that provides the ratings currency for print media, the Audit Bureau of Circulations. The ABC has long been a lightning rod for criticism by media buyers for the length of time it takes to audit circulation numbers -- more than a year in many cases.

That means media buyers are making buying decisions on information that's at least six months old -- a letdown in a world where TV gives overnight ratings and the Web can yield data in real time.

The ABC says it has worked to bring audit timings down to between six and 10 months, but as executive vice president Mike Moran points out, the company still counts every copy by hand, unlike Nielsen or Arbitron, which relies on projections derived from samples.

The main limitation on turnaround time is getting sales data back from distributors of both magazines and newspapers. Another limitation is cost. The manpower needed for manual counts is costly. Faster counts would require even more personnel.

And that's assuming things go well. When there's a concerted effort to deceive the ABC, the system breaks down.

Gruner+Jahr admitted last year that it submitted inflated circulation figures for Rosie, YM and Fast Company. This year, Tribune Co.'s Newsday and Hoy confessed to delivering inflated numbers to the ABC, as did Hollinger Intl.'s Chicago Sun-Times and Belo's Dallas Morning News.

ABC figures, once considered slow but accurate, now have become suspect. Even that company admits that in cases where deception is coordinated and clever, catching fraud can be extremely difficult.

"It certainly complicates it if people are lying and records are being altered," Moran says. Audits caught fraud at Newsday and Hoy, but the Chicago Sun-Times persisted despite numerous audits. "We just don't know what happened at the Sun-Times," he says.

"These things have been very rare in the past," says newspaper analyst John Morton. "I can think of a half-dozen of these over the last 30 years. But I suspect now every publication in America is investigating to be sure similar shenanigans aren't going on in their circulation departments."

Measuring up

Research of all types is subject to error, varying interpretation and matters of opinion. Even the simple matter of counting sales -- recordings through SoundScan, books on bestseller lists, film performance through box office receipts -- can be subject to manipulation or varying interpretations.

But when a ratings point or circulation figure is applied in the marketplace, it becomes currency representing billions of dollars. As audiences segment, each point matters more than ever; fudging it just doesn't cut it any more.

More than at any other time in recent history, the systems with which media are bought and sold are under threat to be replaced, perhaps by an entirely different technology or company.

Fox and Univision are playing hardball. Infinity is betting it can break Arbitron's grip. Print may decide a complete rethink is necessary to stem the yearly loss in advertising market share to TV and the Web.

Says Deutsch's Gardner: "I believe some significant new alternatives are going to happen, enabled by technology -- and it won't be long before we hear of them."