tl;dr: press releases are for bragging rights, Nielsen ratings are for facilitating the buying and selling of advertising.
Last Monday (June 23), ESPN sent out a press release saying that World Cup match between USA and Portugal the day before averaged 18.2 million viewers.
Then the Nielsen numbers came out showing the telecast averaged 13.8 million viewers, nearly a 4.5 million viewer discrepancy. At TV by the Numbers we heard cries that we were reporting the wrong numbers. We’re not, but why are the numbers different?
That’s pretty easy: in the press release ESPN reported numbers for just the match between 6p-8:06p. But the Nielsen numbers for the telecast included the hour of pre-game coverage between 5-6p and adding in that hour drug down the average viewership by nearly 4.5 million.
Some have suggested ESPN is resorting to trickery in their press releases, but that’s pretty silly. It’s the PR folks job to pimp big numbers in press releases. And let’s face it, the general public doesn’t give a crap about how advertising is sold, they just want to see the numbers for the match.
Nielsen’s primary purpose though is to facilitate the buying and selling of advertising.
Indeed, when I asked an ESPN representative why the match-only numbers reported in the press releases are different from what’s in the Nielsen report his reply was, “What ESPN reports to Nielsen is the 150-minute window because that’s how it’s being sold.”
That’s probably good enough for most people, though there is a little more “way-way inside baseball” to the story.
If you’re an advertising supported basic cable network and you run programming that doesn’t contain any ads, while you can have the program measured, the numbers won’t count. If more people watch a presidential state-of-the-union address on NBC than ABC, it doesn’t really matter. No ads, so none of the numbers get counted in weekly or season-to-date ratings.
ESPN certainly wants the World Cup numbers counted in its weekly and season-to-date averages. If ESPN was being dodgy they’d do stuff like try to get the match-only numbers through just on the basis of the halftime ads, or make the pre-game much smaller to beef up the overall number.
Unlike all the major televised sports in the U.S., there’s not a lot of ad inventory with soccer matches and most of the inventory available is in the pre-game. And it makes sense that ESPN would sell the ads in a “if you want to get into the halftime rotation, you gotta buy the pre-game, too.”
That kind of thing happens all the time and isn’t limited to sports (i.e., if you want to buy time on flagship TNT dramas like ‘Rizzoli & Isles’ or ‘Falling Skies’ you have to buy some time on reruns of ‘Bones,’ ‘The Mentalist’ and ‘Law & Order,’ etc.
You’d think that ESPN sells its ads on the C3 rating, so it would not matter that much to advertisers how the programs are broken out in the as-run log. But by artificially lowering the program numbers, they may improve their ad loyalty (ad rating divided by program rating).
I’m sure ad sales are reconciled based on C3. I’m not sure how important the ad loyalty metric is vs. pure C3 average rating for ads, but you’re right that this approach would bolster it!