2013-02-12 07:35:09
Measuring TV audiences keeps getting harder
Feb 9th 2013 | NEW YORK |From the print edition
Babies come from space. And TV ratings are reliable
WHEN it comes to gathering a massive audience, America s Super Bowl shows that television can still whup other media like a linebacker arm-wrestling a mouse. On February 3rd some 109m people (more than a third of Americans) watched the big match on CBS, a broadcast network, according to Nielsen, a research firm. They also watched witty commercials, such as one for Kia cars in which an embarrassed dad, asked about the facts of life, tells his son that babies come from outer space (see picture).
CBS also streamed the Super Bowl on its website to 3m people. But are they different people? How many were watching their iPads and TVs simultaneously? Nielsen, which dominates TV ratings in America, cannot be sure. (In any case, it will not release estimates of online viewing until March.)
If measuring the Super Bowl s audience sounds like a headache, try other programmes. Whereas most people still watch big sporting events live, around half of Americans use digital-video recorders (DVRs) to watch shows later, or order video on demand from their set-top boxes. They are also watching more TV on their computers, tablets and smartphones particularly younger folk. Nielsen s data say that cable and broadcast ratings in America are down by 8% for 18- to 24-year-olds compared with a year earlier. If that is true, advertising dollars could migrate. But young things may simply be watching the same shows on other devices, where viewing is harder to track. No one knows.
Measurement is the number one issue for television right now , says Philippe Dauman, the boss of Viacom, a big media company. Executives battle as hard for Nielsen s ratings as footballers do for Super Bowl rings. These ratings determine where advertisers put the $75 billion they spend on TV in America every year. Recently, however, consumers media-viewing habits have changed too fast for Nielsen to keep up. Everyone is unhappy, says Alan Wurtzel, president of research for NBCUniversal, a media giant. If you can t measure it, you can t sell it.
There are two separate problems with counting couch potatoes, one more pressing than the other. The first has to do with time-shifted viewing, which means that people are watching fewer programmes live. When the DVR became mainstream, advertisers and networks agreed to count eyeballs only if they watched live or within three days of the programme airing. If adverts are skipped by DVR (as around 72% are, according to Bernstein Research), then they are not counted. However, because viewers sometimes watch recorded shows long after they air, media networks protest that three days is too narrow a window. They want to move to live plus seven days. Advertisers, especially those with time-sensitive messages, are not keen.
The second, bigger problem is how to track fragmented audiences. This is a particular worry in America, where people watch TV on countless websites and on multiple devices. Nielsen, a 90-year-old company, had revenues of $5.5 billion in 2011 from measuring the viewing and buying habits of consumers around the world. The largest content companies pay Nielsen $100m-200m a year for its services; advertisers pay it, too.
Who will watch the watchers?
It is a lucrative business, but it is looking old-fashioned. In America Nielsen samples some 22,000 homes that agree to have a meter in their TV, and extrapolates what people are watching nationally. The firm has a separate panel of 200,000 Americans to track online viewing. But critics point out that online viewing is not included in Nielson s main ratings yet, and tablets won t be until 2014. Nor has the firm begun to gather ratings for smartphone viewing, a small but growing trend.
There are so many websites and devices for viewing content that any measurer needs a much larger sample size than 22,000, says Jane Clarke of the Coalition for Innovative Media Measurement, an industry body pushing for new research techniques. In an ideal world one would measure how different people in the same household are consuming media across various platforms, rather than taking a separate data set for non-traditional platforms and extrapolating, as happens now. But it would be tricky to find a fair sample. How many people will volunteer to have their phones, tablets and laptops metered?
More likely Nielsen or someone else will start to use several different panels and incorporate big new streams of data (from millions of set-top boxes, for example). It has been tough going so far, partly because it has been complicated to get a meter to work on Apple s iPads, says Pat McDonough of Nielsen. The minute a technology is invented, [clients] want it measured, she says. But most appreciate it takes time. In December Nielsen announced it would buy Arbitron, a rival with expertise measuring mobile and radio content, for $1.3 billion. This suggests Nielsen wants to buy the new skills it needs, rather than develop them in-house, which would take longer.
Media firms and advertisers say privately they would welcome new competitors to wake Nielsen up. More are popping up, such as comScore, a digital-measurement firm, and Rentrak, which uses set-top boxes to measure audiences. ESPN has announced it is partnering with comScore and Arbitron to perfect its own way of counting viewers across multiple devices. But Nielsen retains a near-monopoly.
And even when there is an alternative source of data, advertisers don t know what to believe. Last September, for example, comScore estimated that Americans spent 4.2 billion hours goggling at online videos. Nielsen estimated that they spent only a quarter of that. For TV firms, which earn a living by promising specific numbers of eyeballs to advertisers, such wild uncertainty is alarming. A happy ending in the form of a reliable measurement standard seems years away.
From the print edition: Business