The Biz: People meter pilots 'need more work'

By Published on .

Most Popular
After a year in the marketplace, two separate local TV ratings' tests-one in Boston and another in Philadelphia-that many believed would yield gains for cable over broadcast, have had mixed results.

Both broadcast and cable advertising associations are claiming some victories in Boston, while in Philadelphia, no one is claiming any gain. For that market, executives say, it's back to the drawing board.

In Boston, Nielsen Media Research has completed a year's worth of data for its new local people-meter system-the remote-control-like device meant to do away with the error-prone handwritten diaries. A similar device has existed for national Nielsen ratings since the late `80s.

Now some local people-meter [LPM] viewer data has surfaced through the Television Bureau of Advertising, a trade association that represents U.S. broadcast stations. The good news, TVB says, is that while cable TV is closing the ratings gap with broadcast, it isn't by much: only 8% to 10%.

For instance, during February 2003, ad-supported cable networks in Boston posted an aggregate six rating points for the adults 25-54 demographic. This was against four rating points in the February 2002 period where only diaries were used. During the same period, broadcast networks scored a collective 25 gross ratings points, down from 28 in February 2002 under the diary system.

"Before LPM [results were] out, many were predicting a major shift in the balance of power between cable and broadcast," said Chris Rohrs, president of the TVB. "We think that the LPM in Boston validates the metered-diaries numbers-that the numbers haven't changed much."

TVB says the numbers are similar across other dayparts, demographics and time periods. In Boston, 35 cable networks take national advertising. But the TVB added in another 20 cable networks-just to be complete and fair with its assessment. TVB acknowledges cable networks themselves in Boston have shown 50% increases in reported ratings year-to-year. But Rohrs says those gains come from a smaller base.

Commenting on the TVB findings, Ira Sussman, VP-research for the Cabletelevision Advertising Bureau, says: "This is a weird number. The aggregated numbers don't make sense and I don't know what it means for buyers." Still, he adds: "The percentages [gained/lost] ring true. Smaller cable networks are underreported. This is absolutely good news for cable."

Advertising executives say these results were somewhat anticipated.

"We expected broadcasting to decline a bit, and cable to be up a bit," said Susan Nathan, senior VP-media knowledge for Universal McCann, New York. "But it's really about other demographics where diaries aren't good. Teens and kids, for instance, are terrible at filling in diaries."

Rohrs says that is an area the TVB doesn't contest with cable-advertising buyers should use cable to get to teens and kids.

`unexplainably high'

In Philadelphia, Arbitron, which had been in the local TV ratings business years ago and now measures national and local radio listenership, has conducted a year-long test of a portable people meter device with Nielsen. The test officially ended in March.

Fifteen-hundred people in its sample carry around a pager-like device for at least eight hours a day, which picks up audio signals-either radio or TV. Because of the portability, the measurement is extended to out-of-home and tracks viewership or listenership in automobiles, stores or bars.

But Arbitron's results are troublesome for a number of advertising, TV and even Arbitron executives. For example, according to the test, network viewership has skyrocketed 34.7%, cable is even higher, up 113.7%, and all TV is 60.5% higher.

"That is unexplainably high," says Universal McCann's Nathan. "That needs to be understood." Says David Marans, senior partner-director of consumer insight at WPP Group's MindShare, New York: "The numbers make no sense." All this data is against historical and current data that show broadcast ratings falling and more modest cable ratings improvements.

More troubling is response rates-the ability to sign up a specific household for the sample-is low, anywhere from 8% to 12%, according to research executives. Typically, Nielsen response rates are in the 30% to 40% range.

Both Arbitron and Nielsen executives say more work is needed.