- The Numbers Racket:
Do not put your faith in what statistics say
until you have carefully considered what they do not say.
William W. Watt
The nets don’t just pull a price out of their respective hats for their advertising time. Advertisers are paying for viewers’ attention. They want to know if they’re getting their money’s worth and the only way to do that is to know how many people are watching. This need has always been so imperative that as early as 1949, just one year after the third network — ABC — went on the air, and even before any of the nets had begun coast-to-coast operations, a regular ratings system was in effect.
Rating broadcast programs did not start with TV. Broadcast programmers had been doing that kind of thing back in the radio days for the same reasons: so that sponsors would know who was watching what, and whether or not those numbers justified the money the sponsor was laying out. A number of TV rating systems have come and gone over the years, but today (and for quite some time, actually), the be-all/end-all of national TV ratings are those supplied by the A.C. Nielsen Company. In the nine decades since “the Nielsens” were first used in radio (the company began measuring TV performance in 1950), they have come to be programmers’ bible of success and failure. While the equipment and methodology of the Nielsens has changed over the years and is often controversial — and continues to change and be controversial — the basic principal remains the same.
The Nielsen company uses a sample of several thousand homes to represent a particular “universe” (the universe is determined by what the client requires: all TVs in the U.S., just cable TVs, just cable subscribers with premium subscription services, etc. To represent the general national audience Nielsen uses a sample of 4,500 homes). The viewing patterns of this universe are measured to determine a “rating” (how many TVs of all the TVs in the specific universe are turned to a particular show) and “share” (how many TVs of just those sets that are actually turned on are tuned to a particular show). Nielsen uses a formula involving the rating, the number of houses using television, or HUT, and the share to establish a particular show’s Nielsen standing and what an appropriate cost for a commercial spot put on the air at a particular time should be. When Nielsen looks at the cost of commercial spots, they talk about the cost per thousand, or cpms for a spot. In English, that means the advertiser’s cost for every thousand viewers watching that show.
Much of the controversy over the ratings has to do with their accuracy, and, more pointedly, exactly what the ratings measure. A Nielsen meter attached to a television set will record that the set was tuned to a particular station at a particular time, but it won’t measure whether or not anybody was actually in the room watching it (there are people who just like the sound of the TV on to keep them company), or whether or not they enjoyed what they were watching (as in, “I would’ve watched something else but everything else sucked even worse!”), and so on.
Over the years, Nielsen, along with the research departments of advertising agencies, networks, and production companies — to name just a few — have tinkered with replacement or additional technologies, but to date nobody’s come up with anything they’re completely happy with.
Back in the late 1980s, one of HBO’s senior scheduling execs once shared with me the problems with viewing measurement systems:
We started out just using meters but we wanted to know if people were actually enjoying what they were watching and how much. The meters couldn’t tell us that so we added diaries. Each metered home also got a little diary, and they were supposed to mark off what stations they were watching and when, and then there was a place to mark how much they liked what they were seeing: very much, moderately, not very much, not at all, and so on.
The problem is the families lied when they filled out the diaries. After the first few days of the month, they would get lazy and stop filling out the diary as they went along. Then, at the end of the month, when they were supposed to turn them in, one member of the household — usually the lady of the house — would fill out the diary from memory.
Well, it wasn’t so much memory, as her filling out the diary with what she wished her family’d been watching instead of what they actually watched. For instance, we’d get back a lot of diaries and there was all this PBS programming written in, but when the meter results came back, PBS ran at a near-zero. According to the diaries, no one in the house ever watched any of our Cinemax late-night adult programming, but then the meter readings came back and it turns out that was the most-watched stuff in the house.
We finally quit using the diaries.
“Ratings,” says respected TV producer Gerald Abrams, “are like dipping a dip stick in the ocean. It’s a small sampling, but so far it’s the best they can do.”
Some of the problems programmers have with ratings may have less to do with how accurate the ratings are then with how good they make a programmer look. In the 1980s, when the networks began to steadily lose audience to cable, the nets began complaining that the Nielsen samplings were slanted towards cable homes. Their point was that Nielsen’s 4,500 homes representing the national market contained a higher percentage of cable homes than the percentage nationally. That, they said, gave the nets an artificially lower rating then they should’ve gotten with a sample they considered more fairly representative.
On the other hand, the nets have their own way of artificially inflating ratings during their “sweeps.” During one week in November and another in February, the nets use the ratings from those weeks to set the standard for advertising rates for the season. But what normally happens during a sweeps week is that instead of measuring a typical week of programming, the nets load the week’s schedule with programming “stunts.” Typical stunting includes running one-time specials, pre-empting a normal night’s schedule to run back-to-back episodes of a top-rated series, airing a top-of-the-line sporting event, and so on.
One curious aspect of ratings is that they rarely have absolute value. When everybody in America tunes into one show, that’s going to be a great rating no matter how you interpret it. But, most programs don’t have that kind of unquestionable standing.
On nights when fewer people watch TV, a show can have a low rating but a high share (in English: a big hunk of a small pie) and be the Number One program of its night. On another night, one where more people are watching TV, the rating of a particular program might be higher but the share lower (a small hunk of a bigger pie). In other words, the same show that pistol-whipped Last Resort on one night can get steamrollered by Big Bang Theory on another night, even if holds the same rating because its share keeps changing. That’s why it’s just as important to schedule a show with as much consideration as it’s produced. More than one good series has been killed by putting it on a bad night at a bad time.
Obviously, ratings are very useful. They tell a network where its strengths and weaknesses are, and it does the same for people buying ad time; vital information when you’re talking about the big bucks that go into TV advertising. The cost of commercial spots can vary enormously depending on the time of day and a given show’s standing in the ratings, but even at the low end, we’re talking significant money. Prime time advertising rates average in the $120-140,000 range for a 30-second spot, but they can range — depending on the show — from low five figures to high six figures. Special events can run the cost of a spot into the rarified atmosphere of millions, with a single 30-second spot during last year’s Super Bowl costing a wallet-killing $2.4 million. All told, the broadcast nets are looking at a whopping $9 billion worth of buys for the upcoming season.
Since the nets’ business is mainly to deliver eyeballs to advertisers, if a show can’t deliver them — at least enough of them to justify the cost of the show — then that show is history. Even if it’s the best thing to show up on the tube since Felix the Cat.
At the end of each season, the networks take a fair bit of flak for cancelling good shows, but, to be fair, the nets have bills to pay. A typical one-hour network drama costs $2-5 million per episode to produce, with a season needing 22-24 episodes. That means, even at the low end, you’re looking at a tab for the season of $44 million — and that’s for a single show. A network needs something in the neighborhood of 21 hours of programming just to fill out its weekly prime time schedule. That makes for big bills!
And while the cost of making things like cars and personal computers usually goes down over time, the cost of making a series paradoxically goes up the longer it runs, at least the above the line costs do (above the line costs would be the talent costs; salaries of the stars. Below the line costs would be the actual cost of production; construction of sets, cost of studio time, etc.). If a show becomes popular, the above the line cost of stars on the show usually rises when it comes time to re-negotiate their contracts. If a performer assumes that they are important to the success of the show, they feel they have enough negotiating leverage to get their salary bumped higher. Some series have been cancelled after several years not because the ratings have fallen so much as it was no longer financially practical to keep the show on the air.
That fiscal burden falls mainly on the producer of the program; not the network. The nets negotiate per-episode fees with producers, but production costs have risen over the years (star salaries aside), and the nets, having lost a large hunk of their audience to cable, have become tighter with their money.
Most network TV programming is now “deficit financed,” which means that the shows cost so much to produce the network fee only partially covers a producer’s costs. Most producers don’t profit on a series until it goes into syndication when re-runs of a series are sold to cable and/or individual TV stations around the country. However, to successfully syndicate a program requires the producer to have enough episodes on hand so that local station and cable network program directors can “strip” the show, meaning they can run at least one episode on at least every weeknight. The rule of thumb is that a producer needs at least three seasons of a one-hour show, or five seasons of a half-hour show on hand to successfully syndicate it (although cable’s bottomless appetite for programming and niche audiences for cult favorites have provided surprisingly successful syndication opportunities for network flops like Firefly, even though the series only ran 14 episodes).
You can see why so many people are so concerned about ratings.
Even the nets would agree, though, that a sad side-effect of the ratings business is that a number of good series’ over the years have been cancelled because the millions of people who watched a particular show represented a number that was millions too low. Real life French Connection cop turned TV producer Sonny Grosso, with over 900 hours of programming under his belt, considers the numbers game and the acclaimed shows that have fallen victim to it, and says, “I’d rather be lucky than good.”
The other rather unappetizing side-effect, in the eyes of many TV critics, has been a network tendency to explore programming whose major value is the number of viewers it can attract rather than its quality as entertainment. The phrase often used to describe this practice is, “appealing to the lowest common denominator,” which is TV programming’s version of throwing chum in the water to draw sharks.
This is hardly a recent trend in network television. It goes back to Day One, something Newton H. Minow recognized when he assumed chairmanship of the Federal Communications Commission back in 1961. Minow, reflecting on where television had gotten itself in just little more than a decade, made a now famous speech to the National Association of Broadcasters in which he said:
When television is good, nothing…is better. But when television is bad, nothing is worse….sit down in front of your television set when your station goes on the air and…keep your eyes glued to that set until the station signs off. I can assure you that you will observe a vast wasteland. You will see a procession of game shows, violence, audience participation shows, formula comedies about totally unbelievable families, blood and thunder, mayhem, violence, sadism, murder, western badmen, western good men, private eyes, gangsters, more violence, and cartoons. And endlessly, commercials — many screaming, cajoling, and offending…
There hasn’t been a TV season since where every TV critic in the country, at least once, hasn’t felt obligated to quote Mr. Minow’s speech.
NEXT WEEK: The Wasteland.