Harder and harder to measure TV viewership (Update)
Every Tuesday, the Nielsen company publishes a popularity ranking of broadcast television programs that has served as the industry's report card dating back to when most people had only three networks to choose from.
And every week, that list gets less and less meaningful.
With DVRs, video on demand, game consoles and streaming services, tablets and smartphones, the way people watch television is changing and the industry is struggling to keep on top of it all. Even the idea of "watching television" is in flux. Are you "watching TV" when you stream an episode of "Downton Abbey" on a tablet?
Nielsen, which has long had a virtual monopoly on the audience statistics that drive a multi-billion dollar industry, last week took an important step toward accounting for some of the changes. Starting in September, Nielsen will begin measuring viewership through broadband devices like game consoles for the first time. Right now those numbers go uncounted.
"The ratings are a very one-dimensional look at what is happening," said Alan Wurtzel, top research executive at NBC Universal, "and we now live in a very multi-dimensional world."
Nielsen's weekly rankings count people who watch a broadcast TV show live or on their DVRs that same day through midnight on the West Coast. To be sure, this is still how most people watch television. CBS didn't need anything other than live numbers to know that its new reality show "The Job" was a flop, and canceled it a week ago after two episodes.
Through separate, less publicized rankings, Nielsen can also track how many people see a program on a time-shifted basis. One ranking, which measures live viewership plus those who watch on DVR or video on demand within three days of the original airing, is what the industry uses to set advertising rates. Other rankings measure those who watch within a week, or even within a month.
Those numbers can present a much different picture of a program's popularity.
During the last week of January, for example, ABC's "Modern Family" ranked No. 12 for the week with 10.8 million viewers if you count just the people who watched on Wednesday, Jan. 23. But within seven days, 15.9 million people had seen the episode, enough to make it the third most popular show of the week behind two "American Idol" episodes. Fox's "The Following" finished a modest 15th place initially, but its audience jumped by 45 percent over the next week, enough to lift the show to fourth place.
Meanwhile, almost all of the "60 Minutes" viewing is done live. The CBS newsmagazine dropped from seventh place in the initial rankings to 15th after a week.
The time-shifted viewing can change a network's perception of a show. NBC would have likely canceled "The Office" years ago without this additional audience. "The idea of how many people are watching a program and caring about the show becomes increasingly important, and it is not reflected in the Tuesday report," Wurtzel said.
CBS considers its freshman drama "Elementary" a case of public perception not matching reality. Last fall, the show averaged 9.7 million viewers—respectable, but hardly a sensation. But between video on demand, DVRs and streaming, CBS said an average of 13 million people watched each episode within a month of its airing.
"If the number the press had seen was 13 million instead of 9.7 million, it would have been seen as a huge hit," said David Poltrack, CBS chief researcher.
In a world where people demand information faster and faster, television executives are no different. They want ratings NOW. The problem is, all of the changes in content consumption demand patience. Nielsen's report on how many people watch a show within seven days isn't released until three weeks after a show first airs—a glacial pace.
"We have to basically train the entire industry to no longer look at the fastest information, which is preliminary and not necessarily reflective of what the reality is," Poltrack said.
Nielsen says it regularly discusses how it releases ratings with all of its clients and there's been no consensus on change. Most people watch their favorite shows as quickly as they can, said Pat McDonough, Nielsen senior vice president of insights and analysis.
Each week the average American spends 32 hours and 15 minutes watching live television, according to a Nielsen study issued last month. More than 12 hours is spent either watching time-shifted TV or DVDs, playing on game consoles, surfing the Internet or watching video on computer or mobile devices, the study said.
"The one thing most people don't think about is a lot of the additional viewing is rolling out slowly over time and right now, live plus same day viewing is the best way to measure," she said. "It may not be that way five years from now."
Networks dispute the notion that things are changing slowly, although they are happy that Nielsen will soon be able to estimate how much television is being watched on broadband. There's a limit to the information, though: Nielsen can't yet tell specifically what programs people are watching this way.
Later this year, Nielsen hopes to roll out a pilot program to identify what people are watching on iPads. It's unclear when this technology will be available for other tablet brands or for smartphones.
The company measures some online video streaming and includes it within its time-shifted reports. However, this picture is partial, too. Nielsen can measure streamed programs only if they have the same commercials shown on TV, and not every website does this.
Netflix's release of an entire 13-episode season of the well-reviewed series "House of Cards" on Feb. 1 was a television landmark, evidence that a lot more "television" content is coming from non-traditional sources. Nielsen has no idea how many people have seen "House of Cards," though. Netflix knows. But it won't tell.
People are increasingly spending time catching up on series they've caught on to midstream, the phenomenon known as binge viewing. No one really knows who is spending an evening watching three episodes from the first season of "Homeland" instead of live TV. Nielsen has an oblique way to illustrate that binge viewing is a reality: When AMC's "The Walking Dead" returned from a hiatus on Feb. 10, the 12.3 million people who watched that night was a series record and evidence that it had attracted new fans during a pause in original episodes.
That episode of "The Walking Dead" was the ninth most-watched television show in prime time that week, but it would have taken some investigation to know that. Nielsen ranks broadcast and cable shows separately even though that distinction means little to a younger generation of viewers. TV is TV.
Cable networks are in no hurry to change that because, with the exception of the biggest hits, even relatively unsuccessful broadcast programs get more viewers than cable.
There's a similar dynamic with PBS. The public broadcasting system generally doesn't pay Nielsen to have its programs rated, although it will on special occasions. The 8.2 million people who watched the third-season finale of "Downton Abbey" on Feb. 17 was more than anything seen on ABC, Fox or NBC that night. No one would have known that unless they'd seen a report generated by a PBS press release.
The numbers-crunchers within the industry know all of this.
Nielsen's Tuesday rankings—and the achievement of getting into the week's Top Ten—used to mean the world. Now it's a small part of television's picture.
Copyright 2013 The Associated Press. All rights reserved. This material may not be published, broadcast, rewritten or redistributed.