Understanding ratings is an important part of understanding the business of radio.

Arbitron is the primary ratings service company. It is to radio what the Neilsons are to television.

Arbitron uses census information to determine the survey area. It then uses computers to generate random telephone numbers. The ratings service then contacts people 12+ and asks them to keep a diary for one week during the 12 week survey period. Even though the survey period is 12 weeks long. It represents what is considered an average week. Listening behavior recorded by multiple people over the 12 week period is reported as if only one week had been measured. Households with more than nine persons 12+ are not eligible for diaries. The company gives a monetary compensation to each person who completes and returns a diary.

Arbitron balances each diary by giving it a Persons Per Diary Value or PPDV The persons per diary value is based on demographic considerations. For example: If more females complete and return diaries than males, each female is given a PPDV less than the PPDV given to each male.

Unlike television which uses Households HUT (households using television) as the basic unit of measurement. Radio uses persons PUR (persons using radio).

Arbitron uses several survey areas when measuring radio.

Metro Survey Area: This geography generally corresponds to a government-defined county configuration known as a Metropolitan Statistical Area. It is abbreviated as Metro, MSA or Met.

Total Survey Area: a geography that has been patterned to include those counties that are significantly covered by the stations licensed to the MSA. Abbreviated TSA or Total.

Area of Dominant Influence: Used for television ratings. It is sometimes used for comparison purposes. Has no real meaning in radio ratings. Abbreviated ADI.

Demographics are used to give a profile of the audience. Common groupings:

12-17 18-24 25-34 35-44 45-49 50-54 55-64 and 65+. Demographics can be further delineated by using gender and ethnic characterizations. They are usually broken down into targets, for example Adults 25-54, Men 18-34, Women 25-54 ,etc.

Dayparts: Time periods which represent one or more days of the week. The usual breakdown:

M-F 6AM - 10AM Morning Drive M-F 10AM - 3AM Midday M-F 3PM - 7PM Afternoon Drive M-F 7pm - 12M Evening M-F 6AM - 7PM Daytime M-F 6AM - 10AM + 3PM - 7PM Combined Drive

Audience estimates are presented along three dimensions: Geography, Demography and Daypart. For example a station's share of the ratings is given in this manner. In the MSA a station has a 5.0 share of Adults 18-34, for Morning drive.

The audience is measured in several ways.

Cume: Represents a station's total reach. In this measurement each audience member is counted only once. This represents the number of people who have sampled the radio station. Regardless of whether a listener tuned to the station one time or many times during the course of the week, he or she is counted only one time.

AQH: Represents the number of people listening to a station during any given fifteen minute period. In this measure listeners may be counted many times. If a member of the audience is tuned to a station for at least five minutes, he or she can be counted.

Cume and AQH are better understood by making a comparison to supermarket shopping. For example Mary and John shop at the same supermarket. Mary makes one 15 minute trip to the supermarket during the week and John makes five 15 minute trips to the same market. In the cume John and Mary are each counted one time. In the AQH, Mary is counted one time, John is counted five times.

If the store is interested in how many different people have visited the store during the week, Mary and John are each counted as individuals. If that same store is then interested in how many people are shopping in the store during any given 15 minute period, John is represented for each of his five trips, while Mary is counted for her one trip to the market.

Two areas that confuse people are rating and share. When people say "rating: they are generally talking about "share".

Rating: An estimate expressed as a percent of a universe. It is a percentage of the people in the market who listen to a specific station. Arbitron no longer reports these numbers.

Share: An estimate of the persons using radio in a market (PUR) who are tuned to a particular station. Arbitron reports these numbers.

AQH persons:. An estimate of listeners to a station for any given quarter hour. For display purposes the last two (00) are dropped. For example "93" does not indicate that 93 people are listening, it means that 9,300 people are listening.

Programmers take the raw numbers and put them into several formulas to give an even better example of a station's audience. These numbers are used to compute several things.

Programmers want to know about audience loyalty. To do this they use the numbers to compute Time Spent Listening (TSL) and Turnover. Sales departments and advertisers want to know how cost effective it is to advertise on a station. They use the same numbers to compute Gross Impressions and Cost Per Thousand or CPM.

TSL

The formula for TSL is AQH in Time Period x AQH Audience / Cume audience.

It is computed by multiplying the number of persons listening during the average quarter hour by the number of quarter hours available in the daypart and then dividing that number by the cume

For example if a station has an AQH audience of 8,500 and a cume audience of 3500 for the morning drive, they can compute the TSL. There are 16 quarter hours available in morning drive, the period between 6am and 10am (4 hours x 4[QH per hour] = 16 QH available in the time period). 8,500 x 16 = 13600, 13600/3500 = 3.88. The average listeners tunes in the morning program for 3.88 quarter hours. To find out the figure in hours divide the quarter hours by four. 3.88/4 = .97. The station now knows the average listener tunes in the morning program for just under an hour.

Turnover:

Indexes the rate at which an audience changes or turns over, during a time period. It is computed by dividing cumulative persons by AQH persons. It represents audience loyalty.

Turnover = Cume persons/AQH persons.

For example 3500/8500 = .411. The .411 represents an index of how quickly the audience turns over. The higher the turnover index, the faster the station's audience turns over.

Impressions and Gross Impressions:

These are used by sales departments and advertisers to determine how many people a spot will reach. One impression is the same as the number of AQH persons. To compute gross impressions take the number of AQH persons times the number of spots aired during the daypart.

For example if an advertiser airs five spots during the morning drive and the station's AQH audience for that time period is 8500, the spot will generate 42500, (5 x 8500 = 42500) gross impressions.

Gross impressions are not the same as individual impressions. Although the spot generated 42500 gross impressions, that can not be interpreted to mean that 42500 people heard the spot.

Cost per thousand. (CPM) is another way of asking how much it costs to make 1,000 impressions.

CPM = cost per announcement / AQH persons.

For example if each spot on a station costs the advertiser $60 and the station has 9,300 listeners during each AQH, it costs the advertiser $6.45 to reach each listener ($60/9.3 = $6.45).

This can be compared with a station having 8,500 listeners during each AQH and also charging $60 per spot. On that station it would cost the advertiser $7.05 to reach each listener ($60/8.5 = $7.05).

Arbitron now provides stations with these figures and more in its Arbitrend reports and in "The Book" as it is called in the industry.

References

Dominowski, P., Church, J., Church, T. (1986). Audience ratings: A primer for noncommercial radio stations Radio Research Consortium.

Keith, M. C. and Krause, J. M. (1989). The Radio Station Focal Press: Boston.

O'Donnell, L., Hausman, C. Benoit, P. (1989). Radio station operations: management and employee perspectives. Wadsworth: Belmont, CA.