FUZZY PICTURE

With Billions at Stake, TV Networks Question Ratings Measurements

By Emily Nelson and Sarah Ellison
Staff Reporters of The Wall Street Journal

Television networks have a lot on riding on ratings. But just how reliable are those numbers?

THIS MONTH'S COVER STORY Adjusting the Set TV Faces a Technological Transformation as It Tries to Win Back Viewers

Family Feud Affiliates Fight TV Networks Over Issues of Ownership and Control

Fuzzy Picture With Billions at Stake, TV Networks Question Ratings Measurements

Outside the Lines ESPN's New Strategy Irks the Sports Leagues That Helped It Thrive

That's the subject of a battle between Nielsen Media Research, the only company that measures television audiences, and critics in broadcasting who fault the company's research methods.

Recently, after weeks of double-checking its data and fighting with the television networks, Nielsen notified TV executives in a letter that men age 18 to 34, a demographic group coveted by advertisers, are watching between 8% and 12% less prime-time TV than they did last year, a sharp decline.

Nielsen says it noticed that viewers age 18 to 34 were watching less TV in July during broadcasts of pre-season NFL games. The TV networks started complaining in September when their new shows launched and many didn't get the ratings numbers among young viewers that the networks expected.

The networks insist that Nielsen's data were mistaken. David Poltrack, head of research at CBS, says more testing of Nielsen's data is needed. The Nielsen sample men may be watching less TV, he says, but "these people may not be truly representative" of the TV watching public.

The networks can't say for sure, of course, because they don't measure viewers themselves. They rely entirely on Nielsen data to decide which shows to cancel as well as how much to charge advertisers for commercial time. An estimated $17 billion in advertising on the broadcast networks was spent last year based on viewer tallies compiled by Nielsen.

Lack of Data

The latest Nielsen controversy reflects long-simmering feuds between businesses that rely on sales and marketing data. Magazines, for example, annoy advertisers by often significantly overestimating key elements of their circulation-estimates that are later scaled back by the industry's official auditor.

The lack of definitive sales data spreads beyond media companies and is becoming an increasingly important issue for a number of marketers, especially since it is no longer cost-effective for stores to stock big inventories of consumer products.

For a mass medium such as broadcast television, with billions of dollars in advertising revenue at stake, the Nielsen sample is surprisingly small. Nielsen extrapolates results from a panel of just 5,100 homes to arrive at its audience tallies. There is no competitor to Nielsen, though others have tried over the years and failed, citing the high cost of entry.

Nielsen measures viewing by outfitting homes with a TV set-top box where viewers punch a button to record their viewing. The data are transmitted via a phone line to Nielsen's data-processing center outside Tampa, Fla., for compilation at 2 a.m. Each day, about 80% of the 5,100 homes send Nielsen data, which works out to 4,080 homes. The sample includes 1,485 men age 18 to 34, so even a small change in viewing patterns can be magnified. TV executives rise at 5 a.m. each day to see early Nielsen ratings and, with just a few ratings points separating NBC, CBS, Fox and ABC, the Nielsen numbers become ever more important.

The most-watched numbers are ratings among younger viewers, age 18 to 34. Advertisers pay the networks a premium to reach younger viewers, because their brand loyalties are still forming and they are harder to reach-the median age of most networks is well into the 40s. A 30-second commercial on NBC's "Friends" costs more than one on other shows because "Friends" has an audience median age of 37.

The current dispute between the networks and Nielsen is the latest in a long contentious relationship with periodic flare-ups. In 1990, after Nielsen ratings showed a drop in young female viewers, the networks complained. Nielsen defended its data, but after a few months, viewing levels returned to their regular levels and the network complaints quieted. Still, there was never a clear explanation or resolution. Many network executives expect the same to happen this time around, too.

"Nielsen is a monopoly," says Alan Wurtzel, president of research at NBC. "I don't believe they have the incentive as a business to try to improve."

'Make Good'

The steep drop in viewers, whether real or perceived, could cost the networks, because it lowers the value of their commercial time. When networks sell commercials, they charge depending on the number of eyeballs they can deliver. If the audience comes in below the promised size, the networks must give the advertiser time at reduced rates to compensate, what the industry calls a "make good." In 1990, the fight with Nielsen over the drop in viewers cost the networks $300 million in make-goods.

Nielsen implemented a new technique this fall, giving more weight to responses from younger viewers in order to more closely mirror the U.S. population. Nielsen said it made the change because it is more difficult to get young people to participate in its surveys. But some network executives say the new technique is distorting the measurement of audience viewership. They also note that new technology, such as digital cable and TiVo recorders, makes it harder for Nielsen to wire homes.