PubTV finds ’09 ratings losses not believable
Public TV programmers are scratching their heads — and those of colleagues and consultants within reach — trying to understand rapidly diving (ah-OOO-ga!) audience estimates from Nielsen Media Research.
Since January, public TV stations in two-thirds of metered markets have seen drops of 20 to 30 percent compared with the same period a year before, says Craig Reed, director of audience analysis for TRAC Media Services, an Arizona-based firm that analyzes ratings and retails Nielsen data to public TV stations. Except for Univision, commercial networks haven’t seen declines as large, he says.
In Houston, KUHT suffered an 84 percent loss of viewers between May 2008 and May 2009, Reed said. Stations in Phoenix, Baltimore and Memphis, Tenn., also confirm that they’ve seen unexplained declines.
The audience losses came on top of losses of 20 to 25 percent registered when Nielsen switched from household meters to people meters in some major markets several years ago—plus the long-term decline felt by all broadcasters as cable has introduced new viewing options.
While some recent changes can be explained by the turmoil accompanying the final major steps in the switch to digital TV, others point to problems in the systems used by Nielsen to sample the viewing.
“I have never in my career seen when you run an analysis, and the audience in one category drops 60 percent, and no other categories are changed,” says Reed. The category of viewers in question is the over-the-air contingent.
Of course, this could have something to do with the historic switch-off of all remaining analog signals from full-power stations. Almost half of public TV’s analog transmitters were switched off between February and June — and the last 25 percent of them on the last day, June 12.
Nielsen argues that many stations are seeing no ratings plunge at all. Despite the analog turnoff, the share of viewing tallied for the PBS-member stations that made the switch June 12 grew by 3 percent after that day, says Pat McDonough, senior v.p. for planning and analysis at Nielsen.
Those stations’ average share was up from 1.2 to a hair above 1.2, she said.
Where stations suffered declines, McDonough told Current, she suspects it’s a temporary dip while the last of America’s procrastinating viewers finally hook up DTV converter boxes and antennas or switch to cable or satellite services.
As of June 12, 2.2 percent of households weren’t ready for DTV, according to Nielsen’s surveys, she says. By the end of the month, that portion was down to 1.5 percent.
(Incidentally, she says, 77 percent of television’s over-the-air viewers have stayed OTA after switching to DTV. But 19 percent went to cable and 4 percent to satellite TV.)
McDonough speculates that public TV’s worst declines occurred in cities where stations were switching their digital channel numbers, especially where they moved from a UHF channel to a VHF channel that requires a different kind of receiving antenna. There the stations risked losing more viewers who were already hooked up for DTV but overlooked the channel change.
In Phoenix, when pubTV station KAET turned off analog a little early, April 29, it switched its digital transmitter from Channel 29 to Channel 8.
Figures prepared by Nielsen for KAET indicate that the drop was concentrated among over-the-air viewers—the ones who had to be concerned with DTV readiness. The station lost 70 percent of its OTA viewers between April and May, says Nancy Southgate, program manager at KAET.
And KAET’s channel change could have contributed. After the April 29 analog shutoff, the station was swamped with calls, she says, and 95 percent of them were viewers who had been watching DTV but didn’t rescan the airwaves with their converter boxes to pick up KAET’s new frequency. KAET may have caught the brunt of this problem because it was the first of four stations in town to relocate on VHF channels.
The analog shutoff is definitely a big problem, but not the only one, says Southgate. KAET saw bigger audience declines than the DTV transition can explain. Even if the station had lost all of its over-the-air audience, it would have lost only 10 to 15 percent, she says.
Comparing a week in April before KAET’s April 29 analog shutoff to a week in June well after, Southgate found KAET’s audience had dropped about 35 percent, from 1.2 to 0.7, she says.
The mystery called Beaumont
The analog turnoff commotion, occurring mostly between April and June, doesn’t seem to explain the large year-to-year declines that some stations began seeing last winter, long before the turnoff.
In December and January, KUHT began seeing declines in gross-rating-points of 30 percent, compared to the same month in 2008, says programmer Ken Lawrence. Like the Phoenix station, KUHT changed channels on its DTV transmitter. KUHT moved to the Channel 8 it long had used for analog.
Though KUHT was waiting until June 12 to switch off analog, it had already lost 65 percent of its audience in May, compared to May 2008. The local people meters and the present mysteries had already knocked the station’s average primetime rating down to 0.24, compared to 1.3 or more in 2007. Viewership was down even for programs with ordinarily stable audience size, such as the NewsHour.
What most startled Lawrence were details of viewing in cities just outside the Houston designated market area. The much smaller city of Beaumont, which ordinarily contributed only 1 or 2 percent of KUHT’s audience, was judged responsible for 9 percent in May.
Station programmers and Reed wonder whether glitches explain the low numbers. Are Nielsen’s sample households representative of the viewers? Are its computers picking up the station IDs inserted in the DTV signals by special encoders? Do those IDs survive the retransmission by cable systems and satellite TV providers?
When the computers have problems picking up those IDs, Nielsen turns to a backup system that identifies stations by matching samples of their audio tracks, Reed says, but is this backup failing because of PBS’s sound synchronization problems?
And what about the inherent shakiness of projecting audience size for channels that get just 1 or 2 percent of viewers at a time?
Nielsen uses a sample of fewer than 500 homes to project the viewing of 674,000 households in Memphis, says Debi Robertson, who schedules pubTV station WKNO.
With a handful of homes representing the entire public TV audience, one or two stray households in the Nielsen sample could have a major effect on a station’s ratings.
This is not why Robertson’s station is canceling Nielsen service, however. They’re doing so reluctantly, at a pivotal moment when she’d really like to know what’s happening with WKNO’s audience. The station is ending its ratings contract to cut tens of thousands of dollars from annual spending.
Web page posted July 6, 2009
Copyright 2009 by Current LLC
