Wednesday, September 4, 2013

Intro to Nielsen Ratings: Demos and DVRs

This week, I'm unveiling a project I've been picking away at for months, an expanded version of my old Intro to Nielsen Ratings post. The three-part series began with the most basic factual info on how Nielsen ratings work and definitions of common terms. Today, we plow forward with the reasoning behind this site's reliance on Live + Same Day and adults 18-49 ratings.

As on the old post, please let me know if there's anything I should add, clarify or correct in these posts. I just want to get everything right.

Popularity vs. Successfulness

Before getting into the nuts and bolts of which numbers I use, a quick word about why I use which numbers I use.

TV ratings are commonly used in the media as a measure of popularity, in order to describe the American culture. "Ten million people watched Jersey Shore, and I weep for the human race!" They don't do a terrible job at this in a broad sense, but attempts to use Nielsen ratings to describe popularity counts are never totally comprehensive, since large segments of the viewing population aren't included. Nielsen doesn't tell us who's streaming programs on Hulu and network websites, who's buying episodes from iTunes or Amazon, or who's illegally downloading.

People tend to hold these things against Nielsen, as if they don't capture the full range of viewing because of a "flaw" in their methodology. But it is worth remembering that while ratings might be used as a popularity measure, they weren't designed as an all-encompassing popularity measure. They were designed to measure specifically the audience for TV commercials, and thus to set the amount of money paid for those commercials. The advertising or lack thereof in other settings is a different beast.

Because the amount of money paid for commercials is still overwhelmingly what makes the TV world go round, Nielsen ratings (or at least the right kind of Nielsen ratings) can be used to give TV industry outsiders a strong window into decision-making.

So this site tends to take the "successfulness" approach to TV ratings before it takes the "popularity" one. "Successfulness" is a much more concrete way to put this stuff to use. Since decisions are made based on shows' relative profitability, we can illuminate a huge part of that process by sticking to numbers that correlate the best with ad dollars. We can get at least a vague sense of whether a show will stick around, whether it will get cancelled, or how it will get scheduled.

Why am I saying all this? The "successfulness" approach helps explain what numbers I use around here. I'm always trying to boil the crowded landscape of data down to what gets to the essence of how successful a show is. I'm not saying popularity or being "most-watched" doesn't matter; it's just not my primary approach. I want to know how the industry works, so I want the numbers the industry cares about reaching.

DVR Numbers

In the mid-2000s, Nielsen addressed the growing importance of digital video recorders (DVRs) in TV viewing by releasing several different streams based on how long DVR viewing was done after the original air date. While pre-DVR ratings only tracked live viewing, there are now streams for live viewing, for live viewing plus DVR viewing on the same day (Live+SD), for live viewing plus DVR viewing up to three days after the air date (Live+3), and for live viewing plus DVR viewing up to seven days after the air date (Live+7). The "overnight" measurement, and the measurement used most of the time on this blog, is Live + Same Day.

While DVRs have been problematic for the networks' ultimate goal, which is getting people to watch advertisements, there is one upside from the network perspective: all these streams of data have allowed the networks to muddy the waters. They often note that since there's so much viewing done on DVR after the airdate, we can't really judge success or failure based on the Live+SD numbers we see the morning after. We've got to wait for the DVR numbers before deeming something a failure. People are very willing to buy into this when convenient, as the anti-Live+SD sentiment is regularly echoed as some way to try to qualify a poor performance by a beloved show on a Live+SD basis.

So why do I continue to use almost exclusively Live+SD numbers on this blog?

The short answer is that Live+SD numbers continue to show a very strong correlation with broadcast advertising rates and with network decision-making. Live+7 gives us a better sense of a show's true popularity, and it makes the numbers bigger (which networks always like). Maybe a bigger Live+7 show will perform better in some ancillary avenues (for example, selling more DVDs). But there's no indication that makes enough difference to make the Live+7 numbers better at evaluating successfulness.

The other short answer: yeah, there's a lot of DVR viewing beyond the airdate, but there's not nearly as much commercial viewing on DVRs. Many people watching via DVR skip commercials, and commercial viewing is what matters, so Same Day ratings will pick up most of the viewing that matters.

And the long answer: while there's a great deal of debate about how much same day and seven day ratings "matter," we actually know what the real stream is for advertisers. The viewing of their ads is what matters to them, so they want ratings that just measure commercial viewing. The one that most "matters" is the one that networks and advertisers have agreed upon as a standard: commercial viewing within three days, also known as C3.

The problem with C3 is that these ratings are very rarely seen in public. So in terms of attempting to analyze the successfulness of TV shows, what we need is the best proxy for C3 within the publicly-released streams. What paints the closest picture of the monetized audience?

From what little we've been able to see of C3, the answer is Live + Same Day ratings. In premiere week 2012, the big four networks' combined rating in Live + Same Day was exactly the same as in C3. And among the few shows whose C3 ratings were released (many of which are consistently among the heaviest-DVRed shows on TV), almost nothing moved by more than 10% in either direction from Live+SD to C3. This general trend continued to hold up in averages across the first few weeks of the season, when nothing differed from the L+SD numbers by more than 7%.

Technically speaking, the difference between Live+SD and C3 ratings is actually the difference between two specific segments: 1) how many people view the program but skip commercials within the same-day window (these people are counted in L+SD but not in C3); and 2) how many people view commercials between same day and three days (these counted in C3 but not in L+SD). Both of these numbers are pretty small; most L+SD viewers are watching live and thus watching commercials, while a great deal of viewers in the three-day segment are skipping commercials. So the difference between these numbers is rarely more than a couple tenths of a rating point.

That's why L+SD is a good proxy. Later streams like Live+3 and Live+7 might do a better job of getting at a show's true popularity, but they include much higher volumes of unmonetized commercial skippers. The networks are technically correct when they say the Live+SD numbers are less important. Live+SD numbers are not the numbers that set the ad rates. But they're still vital for those of us outside of the industry, because they do the best available job of representing the numbers.

There's talk about a move from C3 to C7, which would add another four days of DVR viewing to the mix, but comparisons of three-day and seven-day ratings suggest that this probably gets more press than it deserves given the money. It would only provide an audience bump in the low single digits percent; again, that's because there's not nearly as much DVR commercial viewing as there is DVR program viewing. A low single digits bump to TV revenue would be an enormous amount of cash, but it would not really alter the relative landscape enough to make or break any individual show or network.

I'm not going to go say never pay any attention to Live+7 ever. TV viewing practices continue to shift, so these things are worth re-evaluating whenever possible. Shows with huge seven-day boosts tend to be the ones that get the C3 boosts (even if they're small), so more DVRing may mean even bigger C3 boosts. But in terms of measuring "successfulness," Live+SD ratings remain our best practical answer for now.

Demo Numbers

Similarly, I use ratings among the adults 18-49 age group on this site because the industry uses that demo as a starting point for setting advertising rates. As with the DVR numbers, I make sure to test this approach whenever possible. Even though Ad Age's excellent "What a Spot Costs" articles are supposed to be very rough estimates, those ad rates consistently post a very strong correlation with Live+SD adults 18-49 original averages, while total viewership averages consistently have much weaker correlations. Adults 18-49 ratings are the number that best represent "successfulness" in the broadcast TV world, even for the networks that often try to deny or downplay it.

The industry's adults 18-49 standard is much more controversial than its three-day commercial viewing standard. Why? It seems ageist. Once you turn 50 or 55, you're completely irrelevant! I know all these old people who are swayed by ads all the time! Why don't advertisers care about them?!

So there's a "the way it should be" argument and a "the way it is" argument, not that the latter is really much of an argument. I use adults 18-49 because that's the way it is. Why not write more stuff on how awful the ageism of the industry is? Well... there's really nothing to write about on a day-to-day basis. It's always said that young viewers are so valuable because they're susceptible to advertising and far tougher to reach on TV than the older set. But until we actually see info on the ROI of TV ads, there's no way to get any deeper than that. We can't see if the advertisers are missing out in a business sense by ignoring older viewers. All we ever see is how much they pay for spots, and that tells us pretty clearly that they are ignoring them.

I do tend to disagree with the notion that everything other than adults 18-49 is totally "meaningless." That's not to say we should "account" for every demo equally; the practical uses of numbers like total viewership are relatively minor and secondary. For example, a younger-skewing show tends to be more advertiser-friendly. It'll usually score more bucks per 18-49 point than an older-skewing show. But the older-skewing shows often have the advantage of resiliency, because their audiences tend to watch them more in the traditional ways.

Another quick note on demos: the importance of the various demographics is not quite as clear-cut on cable networks, because I've never really seen any ad rates info on cable shows. Networks often claim to "target" demographics other than basic adults 18-49. And then there are the premium networks, whose bottom lines are about subscribers rather than advertising audiences. Still, I tend to use adults 18-49 with cable as well, just as sort of a generic catch-all young adult measurement that puts everything on a level playing field. But be advised that not all cable 18-49 ratings may correlate as strongly with the ad dollars.

No comments:

Post a Comment

© 2009-2022. All Rights Reserved.