Spring brings the debut of epically expensive made-for-TV movies. Last May, NBC gave us The Odyssey ($40 million). On Sunday, the same network will air Merlin ($30 million). These television events are part of the hype that is sweeps. During the sweep months (February, late April-early May, and November), network execs pack their prime time lineups with the shows they are convinced will reap the highest ratings of the year. Inflated sweep months ratings are, in turn, used to set advertising rates for the rest of the rerun-riddled year. Are sweeps as big a scam as they seem? And what exactly are ratings?
To track viewership, networks (both broadcast and cable) regularly keep tabs on three statistics: 1) Ratings, the percentage of the total households with TV sets tuned in to a show. (Because there are about 98 million households with sets, reports dutifully note that each “ratings point” is equal to 980,000 viewers. Cable networks receive another set of figures based on the number of households wired to receive them.) 2) Shares, the percentage of sets turned on that are tuned in to a show. 3) Demos (short for demographics), numbers breaking out the age and sex of viewers.
S even broadcast networks and 44 cable networks pay substantial annual fees to the A.C. Nielsen company to have these numbers tracked. Every morning, Nielsen compiles the “overnights“–the ratings, shares, and demos for the previous night’s programs, extrapolated from results sent in by the People Meters wired to 5,000 TV sets. (These households are randomly selected but supposedly represent a cross-section of viewers.) Every 15 minutes the People Meters, small black boxes containing a tiny computer linked to a modem, register whether a set is turned on and, if so, the channel to which it is tuned. Attached to the box is a remote-controllike device with buttons that viewers push to register their age and sex each time they watch television.
There is a reason only 5,000 people consent to take part in this program–it’s a pain. It requires that your home get wired–another ugly box in your living room, holes drilled in walls–and that you constantly push buttons. But as a reward, Nielsen families get to pick a prize from a catalogue or receive a $25 a month stipend. The incentives apparently work. Nielsen estimates that 4,500 of the selected households provide data each night. Worried about viewer fatigue, Nielsen has ruled that households participate for only two years.
T he People Meters don’t provide statistically valid data on most (173 out of 211) TV markets. Thus, the need for sweeps. During sweep months, Nielsen sends out “diaries” to 250,000 households across all these markets. In these diaries, viewers log entries every 15 minutes for what they watch. The affiliates use these numbers to set ad rates. (National networks sell ads based on averages of the overnight numbers.)
Through the mid-’80s, networks tied ad prices strictly to ratings. But with the rise of cable and the proliferation of networks, networks lost their leverage to control prices. Ratings now only provide networks a jumping-off point in their negotiations with ad agencies. In addition, advertising has grown more sophisticated, targeting specific demographic groups. This can create unexpected results. CBS, for example, frequently wins the highest ratings. But because it has more viewers who are senior citizens–supposedly the stingiest demo–CBS brings in lower advertising revenue and frequently runs in the red. On the other hand, the WB and UPN networks, with abysmal ratings, are considered comers because of their success with youth–a prized demo.
Many network execs grouch that the Nielsens aren’t reliable. Some networks have flirted with funding a service to compete with Nielsen. The biggest beefs are that the Nielsen numbers: 1) depend on people’s honesty and initiative–filling out diaries, punching buttons, etc. (read Slate’s “My Life as a Nielsen Family” for the confessions of one anonymous fudger of a Nielsen diary); 2) don’t cover a truly representative sample (college dorms, prisons, and African-American homes are all said to be either underrepresented or excluded from the survey); 3) fluctuate wildly–Fox officials, for instance, gripe that ratings for Beverly Hills 90210 yo-yo as much as 86 percent week to week; and 4) measure 15-minute intervals, which is detrimental to networks such as CNN that viewers usually watch for even smaller chunks of time.
The networks’ sweeps tactics, however, take even more heat. Critics say that sweeps cause networks to expend the bulk of their resources pulling out all the stops for these three months rather than spreading them around the year. The result: The cash-strapped networks are forced to air too many blooper shows and reruns in nonsweep months, turning people off network TV. Others complain that during sweeps, local affiliates compete by injecting even more carnivalesque antics into local news. Stunts pulled include: cash giveaways, profiles of the Nielsen families, and tabloid lead stories. And advertisers, of course, complain that sweep months artificially inflate the amounts they pay for spots.
So, if everyone’s unhappy with the system, why does it persist? Some TV execs say their industry is reactionary. Even if the Nielsens are inaccurate–the stated margin of error hovers between 2 percent and 4 percent, but some say it’s larger–the industry will never muster the will to replace them. And Nielsen, which now enjoys a monopoly on the TV ratings business–its major competitors, Arbitron and ARB, having been beaten out of the ratings business–has little incentive to innovate. Its technology is supposedly a decade behind devices used in Norway.