Your doorway to everything genre

@AirlockalphaNo twitter items loaded at the moment ...


SciFriday: Ratings Are Far From Meaningless

Should 25,000 people decide what the rest of the world watches? Yes

Every so often, I find myself using the SciFriday platform to defend television ratings. It seems that efforts to measure television audiences from companies (well, company) like The Nielsen Co. is so hated, that I am almost a lone voice in support of Nielsen.

And no, I’ve never received a dime from Nielsen. Back in 2002, I received a nice email from Nielsen for my efforts in trying to educate readers how ratings work. I’ve also toured the operational headquarters of Nielsen, but only because my day job at the time covered Nielsen as a company, and the place where numbers are aggregated is located just a few miles down the road from Airlock Alpha World Headquarters (all 1,200 square feet of it).

I will never, ever say that Nielsen is perfect. I may never say that Nielsen kicks out new ways to count audiences as fast the technology is developed. But one thing I will say is that for its efforts in counting millions of people watching hundreds of channels each and every day — sometimes down to 15-minute blocks — they do a pretty bang-up job.

So I take some issue when people just slam Nielsen because they blame them for their favorite show’s cancellation (that’s like blaming the New York Times for World War II just because it ran a big headline saying war was declared). Or they think Nielsen doesn’t understand other forms of television distribution like iTunes and Hulu. Or they simply cannot figure out how monitoring the activity of 25,000 people can determine what tens of millions are watching.

And I’ll be honest, those are very fair arguments. Well, not the “kill the messenger” part, but the other two. And I just want to talk about them here, starting first with how 25,000 people can rule the world.

So let’s look at that number. It doesn’t sound like a lot, 25,000 Nielsen families. If you agree with me that there are 110 million television households in America, that means each family represents 4,400 other families. That’s a lot for one family to represent.

But then again, each member of Congress represents an average of 700,000 people. Think about that the next time a tax bill hits the House floor.

Of course, Congress might not be the greatest example of representation in action. However, there is one political method that could help boost the case for Nielsen: election polls.

Every four years, companies like Gallup try to gauge the direction voters are going in by polling random people on who they are going to vote for president. They then crunch those numbers, and share them with news outlets and the American people on a regular basis.

Gallup keeps a close eye on its accuracy. In 2008, its final poll was off 2 percent both ways in predicting that Barack Obama would beat John McCain. In 2004, it was off by just 1.7 percent in George W. Bush’s totals, and even less — 0.7 percent on John Kerry’s support.

Gallup’s biggest gap was 6.8 percent both ways, although it was still able to predict the winner. And that winner wasn’t George H.W. Bush, or Ronald Reagan, or Jimmy Carter, or Richard Nixon. In fact, you’d have to go back even further — that poll predicted that Republican Alf Landon would lose to Franklin D. Roosevelt, in 1936.

Having a 2 percent margin of error in these poll results is pretty amazing, considering Gallup is trying to figure out what about 150 million registered voters are thinking (including many of those won’t even visit the voting booth).

Want to know something more amazing? Guess how many people are surveyed for a typical Gallup presidential poll. Would you say 10 million? How about 1 million? Maybe 700,000, to give ratios similar to Congress? OK, fine, let’s do 25,000 like Nielsen families.

Nope. None of those. In fact, the number of people Gallup surveys in a typical presidential poll: 1,000 people. Yes, each person polled by Gallup represents 150,000 people. And it has an average accuracy rate averaging around 1.6 percent. Sure, Gallup is not always right — it did predict Dewey over Truman, but all polls did that. But for the most part, Gallup is pretty spot on, and it does it with just 1,000 people.

So if one person can predict what 150,000 people would do, why can’t one family predict what 4,400 will do? Random sample statistical analysis is a time-proven method to gauge the public interest, and Nielsen has been using it almost from the beginning of television.

Of course, television has changed a lot since the beginning. To the point where we don’t even need our televisions anymore to watch it. We have iTunes. We have Hulu. We have online players offered by the networks. So many ways are available now to watch content, why isn’t Nielsen spending more time counting that?

The answer to that is far more simple than the numbers game: alternate forms of distribution do not generate the same level of revenue as a live, commercial-filled broadcast.

When you look at network television numbers, it’s not the number of people who are watching “Terra Nova” that Fox and its advertisers care about. What they want to know is how many people are watching the commercials.

Networks pay millions of dollars for ratings not because they want to see who can pee the furthest. They do it because it’s part of their overall business model in selling advertising, by being able to audit the number of people who viewed a commercial.

Advertising rates are based on the number of people who will see a commercial. Of course, those numbers tie back into the number of people watching a show, because outside of the Super Bowl, people tend to watch shows because of the shows, not the commercials.

If you DVR the show and watch it later, the impact isn’t as high, because chances are, you’re going to fast forward through the commercials. Where is the value in that for the advertiser? Nielsen did start to track DVR ratings, but only because it helped networks determine potential audience for a show by calculating an audience base.

So 10 million people might watch “Terra Nova” live, but if 5 million more DVR it, then that means there is the potential of 15 million people who would watch it live on a given night. Not exactly numbers you can take to the bank and cash in, but at least it gives networks and advertisers an idea of the audience potential.

The problem gets worse for iTunes, Hulu and other non-traditional distribution methods. If 2 million people watch “House” on iTunes, for example, Fox should not only be counting the viewers, but counting the $2.99 it collected from each viewer.

Sorry, but no. First, Fox could count these people as potential viewers — like they do with DVR. But since “House” is not produced by Fox, the production company — owned by NBC Universal — will take home the profit from the iTunes fee, not Fox, depending now what kind of distribution agreement Fox has with NBCU.

That doesn’t mean that NBCU couldn’t look at the money it makes on iTunes, consider the future money it could make in DVD box set sales, and then lower its fee to Fox to help keep the show on the air longer. But it also means that those numbers can’t carry the same weight as live viewing does for Nielsen.

Despite what some website bloggers out there will try to tell you, ratings are not the Alpha and the Omega of the television world. While ratings carry a lot of weight, there are hundreds of factors involved in networks deciding whether to continue a show or not. That’s why there are no “magic numbers” or scientific method in predicting whether a show will be renewed or cancelled.

The role of ratings has been reduced quite a bit in the overall equation because of these alternative forms of distribution. But even today, in 2011, networks see their biggest revenue come from advertising sales generated by the people who tune in to a show as it’s being broadcast. That is the model that makes money for the networks, and that is the model that Nielsen, then, concentrates on.

It might not be ideal for the rest of us, especially if we lose a show we like (or have to endure a show we don’t like). But the system is not broken. It’s proven and ever-evolving, but it has to fill the needs of those who pay for the service.

That’s not you. That’s not me. That’s the networks. They believe in it. The advertisers who shell over billions of dollars each year believe in it. And yeah, I’m sorry, that means we have to believe in it too.

This post was created by a person without an author bio.

Could they be a Rut-ro! Shaggy
COMMENTS ARE DISABLED Should we bring them back? Let us know on Twitter and Facebook

Media and Podcast