The Cume

The cume is the percentage of homes in a market that tune to a particular station for six minutes or more during a measurement period.

In an effort to better understand their viewers, public TV stations buy audience data from a commercial measurement service - Nielsen.

With this service comes a myriad of new concepts and jargon, complexities and questions. One concept public TV has employed with particular zeal is the cume (or station reach).

Variables

Several variables affect a station's cume: one of those variables is channel allocation. UHF stations have smaller cumes than VHF stations: UHF signals generally cover less area and reach fewer homes than VHF.

The "market" also affects cume size. Nielsen assigns stations to a single market, determined by where the majority of the viewing in each county is directed. In Nielsen's jargon, this is the "Designated Market Area," or DMA. The DMA is a collection of counties organized around a cluster of commercial TV stations' signals. If a public TV station's tower is near the commercial stations' towers, directing its signal into the "heart of the market," it generally will have a higher cume than if it were alone at the fringe of the market.

The seven-day diary sample also affects a public TV stations' cume size. This is, in part, because the circadian rhythm of public TV viewing is different than commercial station viewing -- for which the measurement methodologies were developed. Public TV viewing is better measured over a longer time period. Also, diaries are a less accurate gauge of viewership than meters. Thus, a metered four-week cume measures a public TV audience better than the one-week diary measure.

Cume Dynamics

To advance this argument, let's look at the concept of cume and how it operates in an individual market. Table 1 presents cumes from the Chicago 1993 February sweep.

The three network stations have the highest one-week cumes: that is, the highest percentage of households in the market that tune in once a week or more. WGN, a VHF of superstation status, rivals the networks' monthly cume. The Chicago UHF independents have signal problems and reach fewer homes than either the VHF indie or network stations.

The two public stations in the market are WTTW, with a strong VHF signal, and WYCC, a weaker UHF signal which telecasts large amounts of instructional TV, even in the evening. The question is: Do meters help public TV cumes?

Meters do increase a public station's cume, whether UHF or VHF, in both prime time and whole week dayparts.

Markets vary dramatically in how much the meters increase the public TV station's cume. Portland, Minneapolis, Milwaukee and Philadelphia traditionally are least helped in the diary/meter increase. Atlanta, Miami and Houston are helped the most.

What does this mean to a public TV station in Iowa or Oregon? At the least, it means that station's cume is 25 to 29 percent larger than reported by Nielsen diary methodology. Barring signal coverage problems, most 28 percent and get a feel for their "true reach."

Monthly Versus Weekly Cume

In the Chicago market cume table there is a greater discrepancy between the public station's one-week and four-week cumes than between the network affiliates' one and four-week cumes. That's because people watch public TV differently than they watch network affiliates: In general, viewers watch a program or two a month on a public station, rather than returning to the station daily or weekly to watch favorite series, news or specials.

Figures 1 and 2 display the monthly and weekly cumes for the 24 metered markets for which historical data exist. Figure 1 displays the total (whole week) reach of the stations in 28 days versus its total reach in "an average seven day" period. For example, KCET ( a UHF station) reaches 50 percent of the market in a week and 72 percent in a month, while to the north, KQED ( a VHF signal) reaches 64 percent of its market in a week and 84 percent in a month.

The average difference between the weekly and monthly reach is 39 percent. In prime time, the 28-day cume increases on the average by 80 percent. In a month, WCET and WEDU almost double their prime-time cume. KLRN, WETA, and KRMA also have large gains. KAET, KETC, and WTTW, the stations with higher weekly cumes than the others, have less to gain in the rest of the month. (Cume is a non-duplicative measure).

The discrepancy between public TV's one-week and four-week cumes is worth contemplating, although it isn't particularly disconcerting. People come to public TV for specific programming, which they watch regularly, perhaps, but not weekly.

Very few homes view the same show each week, even on commercial TV. In a study of Thursday night NBC's 8 p.m. program audience, only 50 percent of this audience watched the program the second week. Only half of the audience that watched the second week, watched NBC the third week.

In a week, an average "NewsHour with Jim Lehrer" viewing household tunes in to one program, In a month, a viewing household perhaps watches one "Frontline" or one "Wall Street Week." Public TV is not central to most viewers' lives or viewing habits. But neither is ABC, CBS or NBC. Viewing is pivotal; the program watched is... whatever is interesting to the person with the TV remote at the moment.

It is clear that measurement techniques have built-in expectations or biases which affect assessment of public TV's audience. The measurement system was, after all, intended to measure commercial stations' mass audiences and meet the needs of commercial advertisers. Small wonder, perhaps, that public TV, with different viewing patterns and different financial underpinnings, should encounter a few problems in adaptation.

But the truth is that these days many public TV stations are concerned about audience maintenance and audience composition. Most stations would be reluctant to relinquish their Nielsen contract--the only feasible quantitative audience data sources currently available to them. And finally, the cume is also how many of the "public" actually watch in a given week or month.

copyright 1998 Trac Media Services, Inc.