TRAC's Rating Service
Choose the Right Tool

Rating and GRPs

Ratings and gross rating points (GRPs) each have certain advantages that make them valuable analytical tools. Their differences make them suitable for diverse treatment and purposes.

Ratings

To refresh your memory, a rating is the percentage of television households that view a program. If there are 100 homes in your market and two are watching your NOVA telecast, then your station has a two rating. The denominator for ratings is constant for any day or any hour: In this case, it is the local market's number of homes with TV sets. Once a year, at the beginning of the television season, rating services estimate the number of homes and people that reside in each market. The denominator remains the same all season.

Ratings are comparable and additive. A two rating for NOVA in Fargo, N.D., and in New York City means the same thing -- 2 percent of the TV households are viewing. Naturally, a two rating in Fargo represents fewer homes than in New York, but the program is equally popular in both markets. If the program had a four rating in Fargo and a two rating in New York, then the program would be twice as attractive in Fargo, in audience assessment terms.

Ratings are comparable across seasons as well as between markets. If NOVA got a rating of two in 1980 and four in 1990, then the program became twice as popular. While the market may have grown in the decade, the ratings would be comparable because the comparison is between percentages of viewing homes, not between absolute numbers of viewers.

Given their mathematical properties, ratings (unlike shares) can be averaged and can also be added to create GRPs. So, if a public TV station has a prime-time average rating of 3.2, then it accumulates 268.8 GRPs in a week (3.2 rating times 84 quarter-hours in a week of prime time). Another station has a 2.7 average rating and hence 226.8 GRPs. Since the comparison is between comparable periods -- prime time in this case -- then the average rating and the GRPs measure and convey the same thing.

GRPs

But there are times when ratings don't tell the whole story. Assume that you are examining two programs -- a half-hour show, Wall Street Week, and a 120-minute movie package, Channel 13 Movies. Each has exactly the same rating -- a two rating. But in terms of GRPs, the movie attracts four times as much audience as Wall Street Week attracts four gross rating points (two rating points times two quarter-hours), while the movie package has 16 (two times eight quarter-hours).

When a programmer assesses audience potential of programs, GRPs are an important consideration. Consider the five most-viewed programs on WQED in Pittsburgh for the October 1991 rating period.

Table 1
Series Day Rating
Columbus Tuesday 4.89
Columbus Monday 4.33
American Playhouse Sunday 3.75
Austin City Limits Saturday 2.77
Columbus Sunday 2.75

Table 1 is a typical list of the most popular programs for a public TV station when a popular mini-series is stripped. But table 2 is a list of the top five programs for Station X ranked by GRPs:


Table 2
Series Day/Time GRPs % of Viewing
Sesame Street M-F/8:30 a.m. 57 7.10
Sesame Street M-F/noon 49 6.16
Columbus (stacked) 28 3.47
Sesame Street M-F/7 a.m. 24 3.00
Mister Rogers M-F/8 a.m. 24 2.94

None of the "most-popular" programs in terms of ratings make the GRP list. (The Columbus on the GRP list is the Sunday afternoon stack.) What surfaces on GRP lists are strips, not prime-time blockbusters. Since GRPs measure viewing tonnage, one can determine the percentage of viewing a series or program contributes to a station's total viewing by merely dividing. Sesame Street's 57 GRPs could be divided by Station X's total 802 GRPs, yielding 7.1 percent of the station's viewing tonnage.

When program GRP lists are compiled and ranked by the percentage of viewing, a number of factors emerge:

  • At any station, a very small number of series accounts for much of the viewing. At Station X, five series account for 23 percent of the viewing devoted to Station X in an average week. These series comprise about 20 percent of the broadcast schedule. If our list were stretched to include the top 25 series, two-thirds or more of all Station X's viewing would be accounted for.
  • Prime-time program ratings are often large because a high percentage of a market's homes are viewing television during prime time (homes-using-television or HUTS) But GRPs help us see that the amount of time devoted to telecasting a series or program is just as important in measuring the use of television as HUT levels. At the average public TV station, 17 percent of the broadcast schedule is in prime time, while about 33 percent of the audience (GRPs) is attracted by prime time. This means that about two-thirds of the viewing is attracted outside of prime time.

Hence, a rating hit parade of highest rated prime-time programs tells one very little about which programs or series deliver the lion's share of the viewing to a station.

Demographic Slices

Examining Ratings and GRPs for different demographic groups gives broadcasters a great deal of insight into how audiences build.

Logically, homes -- being inanimate objects -- do not watch television. Rather, the homes' residents do the viewing. Decades ago, commercial marketers declared households the units of measurement, and that has carried over to public TV. Development people also measure membership in terms of households rather than individuals, since it is rare to have more than one member per household.

Since the 1960s, the emergence of demographic measures brought person ratings and person GRPs for such groups as men 65-plus. The number of people in a demographic category viewing a program is divided by the entire market's population in that category. For example, if 2,000 women aged 35-49 viewed Masterpiece Theatre and there are 100,000 women of that group living in the market, then the program's rating for 35-49 women in that market is a two. The 35-49 women would have four GRPs for the hour-long program.

Examining ratings and GRPs for different demographic groups gives broadcasters a great deal of insight into how audiences build. For example, it is axiomatic that men tend to watch different kinds of programs than do women or children.

Here are Station Y's top-rated programs for women and men, ages 35-49, in late 1992:

Table 3 Person Ratings for Station Y, Market B

Women 35-49 Men 35-49
Columbus Sun 10.04 Columbus Sun 12.22
Great Performances Wed 9.75 Columbus Mon 8.66
Edge Wed 7.32 Columbus Wed 7.81
American Playhouse Sun 6.43 Columbus Tue 7.30
Mystery! (r) Sun 5.97 Masterpiece Theatre Sun 4.67

The men were consistently attracted by "Columbus: Age of Discovery" episodes, while the women sampled the first day and were then drawn away by other programs. (Note the flow of 35-49 women from Wednesday's contemporary dance on "Great Performances" to the premiere of "Edge." Sunday's "Masterpiece Theatre," highly ranked among the men, was a rather unusual offering for the series -- a BBC import, "Sleepers").

Women 18-34 view more than men; older people view more than younger people; children view the most.

Person GRPs give a different picture. In GRPs, the most-viewed programs for Market B women ages 35-49 were: "Mystery" (Thursday telecast), "MacNeil/Lehrer," "Masterpiece Theatre," the Sunday afternoon movie and "American Playhouse." For 35-49 men, the top GRP performers were: "MacNeil/Lehrer," Monday and Tuesday "Columbus" telecast, the Saturday morning stack of "Columbus" episodes and "Masterpiece Theatre."

Many stations have found GRPs useful in compiling viewing tonnage measures. Here are total GRPs for men and women of different ages in Market B in October, 1991:

Table 4 October Person GRPs, Station Y
Age Women Men
18-34 587 268
35-49 614 668
50+ 1066 893

Women 18-34 view more than men (587 versus 268). Older people view more public TV than younger folks. The children view the most (2,444 GRPs for kids 2-5 and 781 for 6-to-11-year-olds).

The table contains an anomaly that has proved true each sweep during the last decade: 35-49 men outview women of the same age. Public TV, given its science documentaries and other reality programming, has done well in attracting men of this particular age group.

Can you add demographic GRPs? For example, could you add women 18-34 and 35-49 to get 18-49 women GRPs from this table? The answer is no. You cannot add GRPs from different demographic groups because the denominators are different. There are not the same number of 18-34 and 35-to-49-year-old women in the market. You could get 18-49 women's GRPs by finding the number of women of that age in the market and dividing it into the number that watched Station Y.

In summary, ratings function adequately to compare program or dayparts of similar length. GRPs function better when comparing dissimilar dayparts and programs of varying lengths. GRPs have an added feature of taking program length into account to create a measure of viewing tonnage.