What Do the Ratings Mean?

[George WK Newman]

Numbers 101

from tvbythenumbers.com

tvbythenumbers.com/demystifying-the-s

Here are some definitions and explanations directly from Nielsen (PDF) updated for the 2008-2009 television season.

Common Terms:

TV HHs: the Nielsen estimate of the number of television households in the USA

PERS 2+: The Nielsen estimate for the number of people aged 2 and over in the USA

LIVE+SD: The number that watched a program either while it was broadcast OR watched via DVR on the same day the program was broadcast (kind of, see below)

LIVE+7: Same as above, but 7 days rather than 3.

Ratings: Nielsen uses a ratings point system. 1 ratings point = 1% of the Nielsen estimate for the category being measured (more detail below).

***

It is our goal to demystify the #s wherever possible. Here are some helpful numbers to keep in mind.

Nielsen uses a ratings point system. One rating point = 1% of the Nielsen universe. There are multiple Nielsen Universes.

TV HHs= Households with TVs

Pers 2+ -this is the estimate of the total number of people age 2 and over in the US for the 2007-2008 season.

These are the two basic sets of numbers that drive the generally available ratings which Nielsen releases.

Keeping in mind that a ratings point is 1% of the universe, all of the aggregate ratings you see are calculated against these two numbers. Nielsen provides ratings for both TV HHs AND Pers 2+. Since these are two different sets of numbers, the lists can be very confusing. I

If you see a HHs rating of 6.2, that means 6.2% of the households were tuned into that program. The math works like this (based on 2007-2008 Nielsen estimates):

6.2% of 112,800,000 = 6,993,600 TV HHs were tuned in.

While TV HHs are a good proxy for how many people are watching, Nielsen also tracks how many people in each household were watching a show. Sometimes you may see something with a VIEWER or Pers 2+ rating. These numbers are just a percentage of the 286,300,000 people estimate.

At TVbytheNumbers, we strive to simplify as much of this as possible by focusing on total viewers wherever possible. We do this because while there are uses for the other numbers, we believe the most important number to the people producing the content, the advertisers and the fans is the actual number of people who saw the show.

Whenever we have the data available for viewers, we use that and rank our lists accordingly. Typically, the “Top Shows” rankings Nielsen sends out are ranked by TV HHs ratings and not ranked by the total number of viewers. Since we believe that the number of viewers is the most important number, we sort our rankings by viewers, not households. As a result, even though we use the same numbers, our lists may look a little different than other lists you see available.

Different shows have different numbers of people per household watching. It’s not the same for every show. So it’s very possible for one show with less households to have more viewers. It happens all the time. Again, we seek to eliminate that complexity by focusing on viewers and ranking our lists accordingly.

Nielsen further breaks their universe estimates for Persons 2+ into age groups (2-11, 18-34, 25-54, etc.). The particular age group ranking generally available is the 18-49 demographic. For the 2007-2008 season, this equals 131,050,000 (or less than HALF of the total universe of 286,000,000).

Advertisers tend to like the 18-49 and 25-54 brackets. Currently we only have access to data segmented by 18-49.

Has your head popped off from thinking about all of this yet? If not, read on.

Time Shifted Viewing – Program ratings for national sources are produced in three streams of data – Live, Live + Same Day (Live+SD) and Live + 7 Day (Live+7). Time shifted figures account for incremental viewing that takes place with DVRs which are currently in approximately 24.4% of all U.S. TV households. Live + Same Day include viewing during the same broadcast day as the original telecast, with a cut-off of 3:00AM local time when meters transmit daily viewing to Nielsen for processing. Live+ 7 Day ratings include incremental viewing that takes place during the 7 days following a telecast.

To simplify this, if you have Heroes on your DVR that was recorded at 9pm, as long as you watch it prior to 3AM, it will count in the LIVE+SD numbers. So same day isn’t exactly the same day, but they always measure this the same, so it’s consistent. If you don’t watch until 4am (why are you still awake at 4am!?) it will not count in the LIVE+SD numbers but it will count in the LIVE+7 (live usage plus 7 days worth of DVR usage).

Households Using Television (HUT): The percentage of all television households in a survey area with one or more sets in use during a specific time period. The sum of the average ratings for a given time period will sometimes be higher than the HUT number because of households viewing multiple programs at the same time. If a household is watching two programs, it is counted toward each program rating but only once toward a HUT number.

[George WK Newman]

What Will We Concentrate on?

Our ratings will focus on the same daily data released by tvbythenumbers.com

This is what the daily ratings will look like (except with our network added in)

tvbythenumbers.com/2009/06/15/sunday-ratings-abc-nba-finals-game-5-lakers-are-golden/20660#more-20660

For individual shows we will concentrate on total number of viewers, and the 18-49 rating.
For the network's nightly performance as a whole we will also concentrate on those two measures, compared to the real world network competition.

I will do my best to have ratings released some time in the early afternoon, the day following broadcast.

[George WK Newman]

Numbers 102

from tvbythenumbers.com

tvbythenumbers.com/numbers-102

FAQ
Numbers 102

We’ve been at this for a while now, so we’re graduating from Numbers 101, to Numbers 102. For more basic information on the TV ratings system, the earlier Numbers 101 is still available here.

This is just really a frequently asked questions (FAQ) list aimed at responding to many regularly occurring questions and sentiments. This will no doubt change over time. This iteration was written January 21, 2009.

We regularly hear the same sorts of things so it seemed easier to have something to point to rather than write the same responses over and over.

1. Everyone knows Nielsen is an antiquated system that doesn’t truly measure viewing. It doesn’t count Internet streaming and downloads, or even measure out of home viewing (sports bars, for example). They should use set-top box data. My “what’s hot” screen on DirecTV looks very different than the Nielsen ratings!!

Whatever Nielsen’s flaws may be (and we’re sure there are some flaws) both the advertisers and the television networks rely on it. Nielsen’s panel attempts to statistically represent the demographic makeup of the United States. Set-top box data and DirecTV do not measure the larger population, are limited samples, and also do not offer much insight into *who* was watching (wife, husband, son, daughter, etc). Advertisers, want, need and demand specific demographic data that Nielsen provides.

For now it seems like the set-top box data is just used as a method to validate the Nielsen data, which we mostly presume that it does, otherwise we figure we’d constantly hear pot-shots from those selling the set-top box data. But we never do. The Nielsen data is the data that drives television advertising sales, and we see nothing coming down the pike soon to change that.

There are many ways to measure online viewing, the networks can get good data on this themselves from their own web servers, and Nielsen, comScore, Hitwise and others offer measurement services for Internet usage.

The real problem here is that as of this writing, the networks haven’t figured out how to make much money on online viewing. iTunes downloads of specific episodes are much lower than many think. 30,000 downloads of an episode over a week, can, and more often than not does, catapult a show into the top downloads list.

When it comes to streaming, far, far more people do this than download shows on iTunes, but the networks haven’t really figured out how to make money with that yet either. It’s a challenged model when you consider that a one hour show on TV winds up with less than 1/4th the advertising in terms of commercial spots when it is viewed online. Basically, they’d need to charge 4x as much money per viewer to break even with television. Advertisers don’t want to pay 4x (surprise!) and adding more commercial spots to online streams risks alienating viewers.

Another problem is measuring who’s watching. The Nielsen system for television reports very thoroughly on that. Determining the demographic makeup of Internet usage isn’t always easy to do via web logs. Nielsen and comScore can provide such information, but then you’re right back to a panel based system that does not measure everybody.

We can all probably agree that business models are being turned on their heads and that new models must emerge, but right now, as far as what they will be, you should consult your Magic 8 Ball.

As long as the buyers and sellers of advertising rely on the data to make the buy/sell transaction, the Nielsen data will be important.

2. DVR Viewing. This show will have a lot more viewers when a week’s worth of DVR viewing is factored in!

That is true for dozens of shows. Many shows that have already been cancelled had more than a million viewers of the show via DVR. We find the DVR trends very interesting, and it’s another factor that is turning business models on their heads. But, we’re pretty sure of a few things with DVR viewing. If it’s good PR for the network, the networks will use it, but mostly it doesn’t seem to wind up meaning much.

There are three types of viewing Nielsen regularly measures and reports. Live viewing, live plus same day DVR viewing (these are the most commonly reported numbers) and live viewing plus seven days of DVR viewing. Almost all of our overnight and weekly data uses the live plus same day measurement. But we do also report some live plus seven data.

But the advertising is bought and sold based on Nielsen data we almost never see, commonly referred to as C3 or C+3 which stands for live viewing of commercials plus three days worth of commercial viewing on DVRs. That’s right, they actually measure how many are watching the commercials.

So far as we can tell DVR viewing doesn’t meaningfully add to a networks ability to make money, and more people watching shows via DVR isn’t good news for the networks.

As you might expect, most people using DVRs usually avoid commercials. As of this writing our experience has been that the C3 data doesn’t differ significantly from the live viewing, and that the relative performance of shows using the live plus same day DVR viewing is a good proxy for predicting success or failure. Particularly, the data for viewers aged 18 to 49, which brings us to..

::cringes::

3. Why all this focus on 18-34 and 18-49 year olds! It’s the people over 50 who have all the money. Why be ageist?

I know it is counterintuitive for some people, but the focus on the younger age demographics isn’t a function of spending power or ageism, it’s a function of relative availability. It’s much, much easier to reach a 60 year old than a 25 year old via television advertising because on average a higher percentage of 60 year olds watch a higher percentage of television than 25 year olds.

Relatively speaking, the 60 year old is easy to reach and the 25 year old far more scarce. It is the relative scarcity of younger viewers which drives the advertising focus and premiums paid for younger viewers.

Whether this is good, bad, right or wrong, I can’t really say. I can say, at least for now, that’s the way it is.

It’s really not about how old you are or how much money you have to spend. It’s about how easy it is to reach you. The easier it is to reach you, the less advertisers are willing to pay for you, the harder it is to reach you, the more they are willing to pay.

4. How do networks decide which shows to cancel and which shows to keep? Networks always sabotage the shows I like and wind up cancelling them. Why do they cancel good television in order to put on more reality crap!?!?

Bill Gorman developed a relatively simple and straightforward measure for determining the likelihood of a show being renewed or canceled, and while there are definitely some exceptions, thus far it has been highly accurate and a very good tool for predicting. How does it work? Generally speaking, it’s pretty straightforward:

For each network we look at overall prime-time season-to-date averages among 18-49 year old viewers
We then look at the season-to-date average among 18-49 year old viewers for the particular shows and compare them to the network’s average
Shows that perform way above average are kept
Shows that perform way below average are cancelled or not renewed
Shows that are average are kept
Shows that are a bit less than average but not way less than average are the interesting “on the bubble” shows. Some are kept, some are cancelled. It could go either way! It’s those shows that are a bit less than average that are the hardest to predict and seem to wind up drawing the most attention
There is no doubt that viewers often take how networks handle their favorite shows personally. But we do not believe that networks are out to sabotage particular shows. Networks love shows they can make a lot of money on, like shows they can make good money on, and don’t like shows they can’t make much money on.

Relative performance among 18-49 year old viewers makes for a pretty good proxy about a show’s ability to make money. This matters more than overall viewership. That may not seem right to you, or fair to you, but it definitely seems to be the way it has been in the past and continues to be as of early 2009.