Newest Nielsen “science” boosts cable, deflates broadcast

I’ve just spent the past hour studying a document that describes Nielsen’s “Zero Cell” ratings model, the latest gimmick from a company whose statistical analysis has determined broadcast revenue for decades. A friend told me, “You need to be a Navaho Windtalker to decipher this stuff,” and I think that’s fairly accurate. I’m going to try, however, because it’s yet another big problem for local broadcasters.

This has been in place for a couple of ratings periods, but the results are really just beginning to be felt now as local sales professionals hit the streets with numbers that, in some cases, are very different than they were before.

A little history, and please remember that this is MY interpretation: As cable brought about audience fragmentation, this created an enormous problem for Nielsen, because the whole game is based on sample size. Television advertising isn’t sold by household numbers; it’s sold by demographics within those households, so fragmentation is a big deal for statistical analysts, who need big sample sizes to be confident about the numbers.

Nielsen measures audiences three ways today: meters, diaries and local people meters. Despite technology, most demographic measurements (ages, gender, etc.) are determined by the oldest method, the diaries. While meters do a good job of capturing household viewing, very often there’s no quarter-hourly match in the diaries, and that’s a problem when it comes to demographics. Here’s the way the Television Bureau of Advertising (TVB) puts it:

In any single quarter hour when there is meter tuning to a source (indicating HH tuning occurred), but no corresponding entry in a diary (to indicate which demographic was viewing), a zero cell occurs. This is also referred to as “tuning without viewing.”

This has been around since meters were first introduced, but the exploding cable universe led to more and more audience fragmentation, and that led to more and more “zero cells.”

Since the lion’s share of these “zero cells” came from cable viewing, the cable industry bitched and moaned that they were getting screwed. Nielsen apparently didn’t move fast enough, so the industry created its own measurement for demos and called it “Fusion.” The methodology apparently didn’t meet Nielsen or the TVB’s “standards,” but the cable industry didn’t give a crap. This is about money, folks, enormous sums of money.

Sensing on which side their future bread was going to be buttered, Nielsen last year announced this “Zero Cell Ratings Model,” which populates some of those “tuning without viewing” quarter hours with demos from what it calls “donor quarter hours,” or the demographic skew of quarter hours with similar tuning/viewing patterns.

Following me?

The end result is that the latest official Nielsen ratings show that while the homes using television (HUT) level stays the same, the persons using television (PUT) levels have gone way up. This has negatively impacted demographics for broadcast sales but positively impacted the same for the cable industry. Some would argue that the new Nielsen strategy more accurately reflects real viewing, but that’s not the point.

Not only is the universe into which broadcasting casts its net shrinking, but the company that measures that universe has adopted a new method that cuts the pie slices even thinner.

In football parlance, this would be called “piling on,” and it’s just another challenge for the beleaguered television broadcasting industry.

Comments

  1. navaho windtalker says

    um,

    readum numbers… still can’t makes buffalo heads or tails of them.

  2. Of course the predictable math you get from “donor quarter hour” measuring is some very strange projections. While PUTs will continue to fall into some kind of understandable pattern, projections will unravel into anarchy. As you know, ad buyers will take demos and project them into the time of the year they are actually placing a buy. So a buyer will take a February ratings and multiply buy time period HUTS of May because their buy will run in May-June. This is all fine and dandy as long as normal viewing patterns exist. Of course “donor quarter hour” estimate throw normal viewing pattern to wind. Because now you are not only estimating HUTs for an upcoming time of year, but you are also estimating “similar” programming patterns to address the zero cell issue. So let’s see if we get this calculation right: We’re going to estimate HUT’s based on past viewing patterns, using the most recent rating multiplied by those estimated HUT’s, combined with the fact that those ratings are based on not on actual program ratings but “donor quarter hour” estimates of those demos…and of course we all understand that everyone that receives a rating book fills it our carefully and accurately. And you want me to invest how much into the calculation?

Speak Your Mind

*

This site uses Akismet to reduce spam. Learn how your comment data is processed.