There is a wonderful little book that I highly recommend. It is called “How To Lie With Statistics.” It was written in 1954 by Darrell Huff, and it is as relevant today as it was 56 years ago. I read that book about five years ago, dog-eared it, and made notes. It gave me incredible insight into how numbers are generated and how numbers are presented in the form of statistics. It makes it easy to question statistics.
On April 12, Nielsen Research released results of a study on the “Fourth Screen,” designed to bring awareness to the general Out-of-Home audience and the potential for exposure.
In the report, they cite that during the last four months of 2009, “adults 18 years or older were exposed to 237 million video ads per month in the ten location-based video networks included in The Nielsen Company’s inaugural ‘Fourth Screen Network Audience Report.’”
Their breakdown of ad exposures by network is as follows:
Source: Nielsen.com
The press release further states that, “For example, the C3 average audience for a prime time broadcast TV commercial was three million viewers age 18+ in October 2009. During that same period, video ad exposures to NCM’s and Screenvision’s movie theater networks combined for an average 61.7 million, meaning that it took about 20 primetime ads to reach the same audience as a typical month-long advertising flight on both movie theater networks.”
Okay. Fair enough. 61.7 million exposures on two networks in one month. Not too shabby. But let’s snoop around your TV dial, shall we?
I found that the ratings from October 10, 2009 had a total of 68.9 million viewers (ages 18+) over a single night with an average of 4.6 million viewers over 14 broadcast programs. And that was only five networks (CBS, NBC, CW, ABC, FOX). I have no idea if October 10 was the big night in October, but that’s one night, not a whole month. I also have no idea if those viewers were watching commercials.
The C3 measurement is a measurement of the broadcast commercial time plus three days on the DVR (hence the “C3″). So let’s assume that the three million Nielsen states is over a four-day period. Nielsen has no idea if these people are watching the commercials, either. And last time I checked, four days doesn’t equal 31 days.
Time To Get All Murky Up In This
Let’s play with some statistics. Let’s assume those October 10 numbers were consistent and those programs ran four times in a month (or four weeks of four day C3-style exposures). Now you’re looking at 275.6 million expressions in a month, against NCM’s and Screenvision’s collective 61.7 million expressions per month.
Going back to that one night, let’s divide NCM/Screenvision’s total by four. You get 15.1 million, compared to broadcast’s 68.9 million.
Step back for a moment. I said that it was 68.9 million viewers for one night. Perhaps Nielsen is saying that over 31 nights, the average number of exposures was three million a night. And maybe they’re throwing more networks in there, like PBS and QVC and the Disney Channel. Okay. But then that would force me to take the 61.7 million and divide that by 31, right? If so, the nightly number of exposures for NCM and Screenvision would collectively be 1.99 million. That’s over a million less than broadcast. Right? So broadcast wins.
Hey, let’s have some more fun. The MPAA (Motion Picture Association of America) claims there were 39,717 total cinema screens in operation in 2009. According to NCM and Screenvision, they collectively account for 31,800 screens, or about 80% of the total number of screens.
I found 62 additional national networks on my own cable programming. That means that the five networks only make up only 7% of my total TV lineup. And that doesn’t include any HD channels or any premium channels (like HBO). That 7% would shrink to less than 2% if I could add in the extra 261 available channels.
Before you argue that those are cable channels and not over-the-air broadcast, I found 39 channels that claim to have access to at least four million households in the U.S., although some of them are not national. I used the 4 million mark as a middle ground between Nielsen’s 3 million, and my 4.6 million. In this case, the five networks make up 11% of the available broadcast channels.
But those numbers aren’t nearly fun enough. So let’s stick with the cable, shall we? I found those on Wikipedia, and everything we read there is 100% correct, correct?
So let’s quickly recap with my statistics:
61.7 million exposures a month over 80% of the cinema screens in the U.S.
68.9 million exposures in one night over 7% of the broadcast channels in the U.S.
At 275 million expressions a month, that beats the 10 total location-based networks by 38.2 million exposures. So I have proven unquestionably, undoubtedly, indisputably, and clearly, that broadcast is still the best way to expose an audience to a commercial.
Performance. Night after night.
Let’s spin it a little more: Nielsen claims that 114 million homes have at least one TV set. Let’s assume that every single set was on only once per day for one program. You would have 114 million potential views per day, or 3.42 billion potential views per month. By comparison, NCM and Screenvision together see approximately 1.18 billion moviegoers a year, or only about 98 million moviegoers a month.
Keep spinning: Take that 4.6 million viewers from October 10, 2009, and multiply that over 31 days. Now you’re ate 142.6 million viewers. Yep, that’s still more than the movies. Heck, take Nielsen’s “three million” average. In October, that would be 93 million exposures, only six million shy of the movies.
But wait. Hold on! Most television stations run 24/7. Do most cinemas? If that’s the case, then…oh forget it. You get the point.
A Nice Figure Always Has Verbosity.
So let’s bicker about verbiage. Nielsen makes the claim that this is the “Fourth Screen” report, indicating that the Fourth Screen is the Out-of-Home sector. According to Nielsen, the first three screens are TV, Internet, and Mobile. However, most in the digital signage industry seen cinema, TV, and computers as the first three screens. At this point, some believe that mobile is part of the fourth screen of Out-of-Home, while some think mobile is the fourth screen with the rest of OOH making up the 5th screen. Under this believe, cinema doesn’t belong in OOH category, correct? So the chart above would lose that 61.7 million exposures. TV keeps looking better, doesn’t it?
To make matters worse, the paper is highlighted in an article on AdAge on April 14, 2009. The article freely mixes the term “expression” with the term “impression.”
An expression is someone seeing an ad. An impression is that person remembering it. And those two terms are not interchangeable, not at all by measurement standards. If you think they are, do this: Go to the local movie theater and count how many people are there before the commercials and trailers start. Those are the people “exposed” to the commercials. After the movie, ask all of them to make a list of all the commercials they remember seeing. Here is what you will notice: Those who remember seeing ads are “impressions.” Those who were impressed will never be 100% of the audience. And those who were impressed are the true currency through which the value of a network is determined. Oh, those who were impressed are also the most difficult to measure.
To be fair, the Nielsen press release does not use the term “impression.”
This is not an indictment of any network or Nielsen. I’m all for figuring out the right way to measure audiences and determine value. I take full responsibility for butchering the statistics and adding in others to make my case. Under no circumstances does this make me an authority. But I bet I made a convincing argument, or at least confused you enough to question what you are reading from Nielsen. You can play with statistics in any number of ways that benefit your side of the story. And that’s dangerous when talking about an industry that is so nascent with understanding the true value of an impression. It’s easy to rally around numbers. That doesn’t mean the numbers are accurate or, in some cases, real. Context is everything.
Now, go buy “How To Lie With Statistics” and really see the other side of the number. I’m going to take some aspirin because my head hurts. A lot.
