My love/hate affair with NBC's Total Audience Measurement index (TAMi) is still on.
First the reason for the love, because it's easier to explain. We often get questions like: I wonder how iTunes downloads do in comparison to DVR numbers?
They do so badly relative to the television audience, and even the DVR audience, that NBC now simply lumps iTunes, Amazon and -- get this -- cable video on demand (VOD) and mobile services all into one bucket. Even in that combined bucket, it's pretty small. For the Heroes episode that aired on October 27, we know that the average live+7 audience for Heroes was 9.68 million and the DVR viewing component of that (though a lot of it happened the same night the show aire) was 2.73 million.
Combined VOD and Internet downloads for that episode were 68,966. So even DVR viewing is more than 38x as much as their combined category they stuffed downloading into. The overall average TV viewing winds up being 9.68 million to .07 million, and that's with favorable rounding! So at best, even if you assume it was all downloads and no VOD viewing, and that it was all iTunes it represents less than one percent of the viewing (seven tenths of one percent) if you add it to the TV viewing numbers.
Downloading is a rounding error. DVR isn't a rounding error as it represents about 28% of the total viewing.
The primary reason I hate TAMi is that it's really more about PR than about producing something the advertisers could really get their arms around. I'm not shocked, but it's kind of annoying.
If you look at the latest TAMi report you'll see they report a TV audience of 13.549 million for the same episode of Heroes I referenced above that had an average audience of only 9.68 million. Both numbers are from Nielsen? How can it be so!? It's because what advertisers care about is average audience, so that's what's reported here, and pretty much on any trade publication or blog site that tracks ratings. NBC is using the total audience measurement which is Nielsen-ese for "watched at least six minutes."
I know, I know, "But Robert, it's the TOTAL audience measurement index, duh!" But it gets convoluted pretty quickly. I can see the difference the puffery makes with television viewing, but there's no way to get a handle on Internet streaming without seeing the minutes.
NBC includes any stream started as a stream whether you watch the whole thing or 5 seconds. There are a lot of streams, but even if you watch only ONE episode of Heroes on NBC.com, they typically (and this is according to them!) break it up into each segment that was separated by a commercial. So if you watch all of one episode of Heroes on NBC.com, you typically get counted six times. And if I'm only one person, and get counted six times -- that's not a total audience measurement. That's inflation by 6x and pads the Internet stream counts nicely.
Hulu doesn't work the same way, if you watch a full episode there, you only count once. Of course if you only watch 3 seconds, you count once as well. The have over four million streams, but whether that accounts for 500,000 views of the whole episode, a million or two million, we can't know. And since there's a significant difference between 500,000 and the 4.276 million reported, I'd kind of like to know.
Despite my love/hate relationship with TAMi (I also have a love/hate relationship with Heroes, if it means anything to you), I am happy NBC does it, and I'd rather have them produce the convoluted mess that is TAMi than not produce anything at all.
At least until they begin lumping in VOD, mobile and downloads with Internet streaming.
For those who missed the link to the latest TAMi report above, here it is again. It's worth reading all the fine print of how things are counted and some data is delayed before reaching any conclusions about the trends with the most recent episodes of shows.
P.S. some shows, like Chuck, which obviously are available on iTunes and obviously have had some downloads, for whatever reason, still show no downloads.