Explainer

How Does Sweeps Week Work?

Why networks roll out the big guns—and why advertisers put up with it.

A Friends wedding sweeps in the viewers

A much-hyped wedding on Friends, an all-star Survivor, a host of blockbuster movies-of-the-week —it can only mean the return of February sweeps, when networks rely on shameless stunts to get viewers to tune in. Couch potatoes know that broadcasters use the viewership numbers collected during sweeps to set advertising rates for the rest of the year. But how do sweeps work, exactly, and why do advertisers put up with a system that seems contrived to inflate audience numbers and ad rates?

Actually, sweeps week doesn’t set rates for all advertisers—just local ones. The standard half-hour of television contains 22 minutes of programming and eight minutes of advertising, of which six minutes is national and two minutes local. Ad rates for the six minutes of national advertising (typically bought by large companies like Ford and Budweiser) are based on Nielsen ratings, which sample the year-round viewing habits of a small number of “Nielsen families” across the country. Rates for the two minutes of local ad time are based on sweeps, which run for just four weeks four times a year (in February, May, July, and November) and survey 210 local television markets ranging in size from greater New York City’s more than 7 million TV households to the 15,670 in North Platte, Neb.

Sweeps began 50 years ago when the AC Nielsen Co., which dominates the television rating business, mailed out diaries to selected households across the country and asked the residents to record their television viewing habits. Because analyzing so many handwritten diaries was labor-intensive and expensive, Nielsen couldn’t measure viewing habits year-round and decided to extrapolate from four annual surveys. After each survey, the diaries are collected by region, starting in the Northeast and then sweeping—whence the name—across the country. Networks then use the results to set local ad rates for the next three months. 

The savvy networks soon realized that a year’s worth of advertising dollars was riding on 16 weeks of viewing habits, so they began to fill the sweeps weeks with spectacular one-off gimmicks to get viewers to watch. Today, this practice makes sweeps weeks almost perfectly unrepresentative of the year as a whole. Local advertisers pay to reach all the viewers who tuned in for Phoebe’s wedding, but only the people who stick around to watch lackluster episodes in nonsweeps months actually see their ads.

Advertisers sometimes grumble about this bait-and-switch, but sweeps have managed to survive nonetheless. Networks defend sweeps because they fear that actual year-round viewership is lower than the current system suggests; with a more accurate count, they might have to charge less. Advertising agencies and media buyers are paid a percentage of their clients’ total ad spending, so they have an incentive to support a system that inflates ad rates. And the local advertisers—because they are scattered across the country, make small ad buys, and have less clout than their national peers—are stuck because the networks deliver an audience of unparalleled size that they can’t afford to boycott in an attempt to force a change in the system.

Since your TiVo can now track your viewing habits minute-by-minute all year long, it might seem that new technologies will soon spell doom for sweeps. But the system has proved to be a sort of QWERTY keyboard of the media world, in that a standard set by the technological limitations of one era prevails long after those limitations have been removed. The handwritten diaries, of which 1.6 million are edited each year, are still used in most regions, although Nielsen has been slowly rolling out electronic “people meters” that attach to televisions, satellite dishes, etc., and continually monitor what is being watched. On Feb. 4, Nielsen announced a deal with TiVo to receive data on the viewing habits of Americans with digital video recorders (Forrester Research estimates there were 3 million in use in 2003). But for the moment the deal is intended merely to fill a gap in survey coverage rather than provide a substitute for the diaries and people meters Nielsen continue to rely on.

Next question?

Explainer thanks Sheryl Hayward and Mark Cooper.