Printer Friendly

The age of the single source.

The Age of Single Source

In today's marketing environment, advertisers are concerned about the productivity of their advertising and media choices. Single source research enables an advertiser to make more informed media decisions because the research is designed to measure each respondent's demographic and product purchasing characteristics and how magazines, newspaper, radio and television are used by these respondents.

There are two kinds of single source measurements: . High Tech Single Source . Traditional Single Source.

High Tech Single Source measurements will enable an advertiser to evaluate the effect of media advertising, promotion and pricing on product movement. In view of advertisers' demand for accountability as marketing costs increase, High Tech Single Source offers actionable input to the marketing process. Like death and taxes, eventual acceptance of High Tech Single Source measurements is a certainty.

Companies such as Scantrack (A.C. Nielsen), Infoscan (IRI), and Scan America (Control Data) are marketing High Tech Single Source or testing it in a few markets. These companies use meters to measure television viewing and electronic scanning devices to track supermarket product movement. It is certainly possible in the future to expand this approach to cover all other major media.

In the musical Hair, lyrics were sung about "the dawning of the age of Aquarius." Single source is a concept that is well past the dawning, but still in the early morning stage of development. When the Age of Single Source finally arrives, it will enable an advertiser to nationally and locally measure the productivity of pricing, merchandising, promotion, and advertising stimuli to the marketing process. According to plans, however, High Tech Single Source will not be able to accomplish such ambitious goals until the mid to late 90s, and it will be expensive.

On the other hand, there are Traditional Single Source syndicated services like Mediamark Research and Simmons Market Research Bureau available now that are national in scope, have large samples, and can inexpensively measure product use ans consumption, not only by television program but against all media. While the Traditional Single Source research services do not provide continuous daily or weekly product movement, the aggregated date are easier to analyze and can improve the selection of media.

Popular Myths

Despite the fact that the syndicated services have been measuring product/media relationships since the 1960s, agencies and advertisers have generally resisted using anything but the usual demographic profiles for program selection. Program decisions are based on an indirect match of age, sex, and socio-economic ratings. A direct match of product users, or volume consumed by program viewers is rarely used. Formal attempts by CBS-TV to directly match program selection to product consumption, using its "Consumer Audience Profile" (CAP) approach, have not been successful.

Yet, there is mounting evidence that the Traditional Single Source measurements offer an opportunity now to improve the productivity of television advertising purchases. Agencies and advertisers have generally resisted using this research in the past. There are four reasons given by agencies for not using product data in TV program selection: . It's not practical. . Product data as measured by the syndicated services are not reliable. . Product data are not available for all programs and are not predictable. . Selecting programs on the basis of users or volumetrics does not discriminate sufficiently between programs to make the effort worthwhile.

The Issue of Practicality

Is it practical to evaluate TV programs (or any medium) on the basis of product users or volumetrics? It is certainly practical in the planning stage to seek out TV programs or program types that deliver more target customers as measured by viewers' product use or volume consumed (volumetrics). It is also practical to employ product use or volumetrics after the programs have been selected in the corporate buying process. There is ample time to improve the allocation of programs to each of the brands in the corporate stable.

Critics, nevertheless, point out the difficulties in using volumetrics. They refer to the lack of linkage between syndicated sources and network program rating research generated by Nielsen and AGB. While they acknowledge the power and speed of desk top computer applications like MEMRI (Microcomputer Evaluation by Mediamark Research) to evaluate programs on user/volumetrics, they discourage their use. They cite several practical considerations: the time required to learn new systems, manpower priorities, cost of hardware and software, and the fact that age and sex matching which is now used by advertisers and agencies works.

These critics agree, however, that syndicated reports play a major role in planning television advertising. Most often syndicated research is used to profile the product user, and buying specifications are set based on age and sex criteria. Buying commercial time according to the age and sex composition of the audience does work, but there is evidence that is possible to extend media effectiveness for clients using single source volumetrics.

Are the Data Reliable?

Not withstanding the ease of application, are the data produced reliable? There is a belief that syndicated services' TV audience data are not as reliable as measurements based on meters. The program measurements of MRI or SMRB correlate with Nielsen on sex and age. The levels, however, don't always agree since viewing reported by MRI, for example, measures program audiences not average minute audiences.

The prime interest of media buyers is to develop indexes of product user selectively by program. The primary criticism of these data are the variability of program product volumetrics from report to report. The stability of program selectivity is critical since advertisers want to make reliable program decisions.

Do volumetrics by program remain reasonably consistent from syndicated report to syndicated report? MRI produces two studies per year. One is in the Spring; the other is in the Fall. There is virtually no difference in consumption reported for 22 common programs measured both in the Spring and Fall (Spring-110; Fall-112).

There is little variation by program. For example, 15 or more than two-thirds of the program reported consumption in the Fall that was within ten percent of the Spring survey period. All 22 programs were within 15 percent. This stability by program is well within the relative sampling error we normally use to judge the stability of program ratings by household, age or sex.

Critics point out that MRI reports twice yearly (Spring and Fall) with results from the preceding two six-month waves averaged. As a result of this overlapping, the reported data are smoothed. A comparison of reports containing no overlapping interviewing waves was undertaken. Fourteen programs measured in both the Spring 1986 and Spring 1987 MRI reports were compared on the basis of how they were indexed by Cat and Dog Food Consumption. Removing the common interviewing wave does not materially affect the average program index of consumption. The average program reported an index of Cat and Dog Food Consumption of 104 in the Spring 1986 and an index of 100 in the Spring of 1987.

Further evidence of the stability of product user measurements by TV program was reported at the Advertising Research Foundation (ARF) Electronic Media Conference in December 1987. Jim Spaeth, Vice President of Scan America, tested stability by ranking 47 prime time programs by household, women 25-54, and product use ratings (called Buyergraphics). These rankings were then examined over time one quarter and two quarters forward. On the basis of rank order correlation, the findings were summarized, "Whether we're looking one or two quarters forward...the rank order correlation for Buyergraphics are in precisely the same ball park as those obtained for demographics and just slightly lower that those obtained for households."

Examining Predictability

One of the problems in using volumetrics or consumption data is that in a dynamic medium like TV, product data are not measured for all programs. The product consumption of new programs, in particular, is not available at the time of negotiation or even during the corporate allocation stage. The prevalent belief is that unlike age and sex ratings that are readily predicted, product consumption is not predictable by program.

To test our ability to estimate consumption, MRI developed an estimating technique to predict program volumetrics. MRI predicted Dog and Cat Food Consumption for each of 31 programs included in the Spring 1986 report. The technique used was to apply Dog and Cat Food category per capita consumption to the age distribution of each of the 31 programs analyzed. These estimates were then compared to the actual consumption report by MRI in the Spring 1986 study. The results are impressive. The average program's actual consumption (as reported by MRI) was 14.4 million cans used in the last seven days. Based on category per capita estimating procedure, the average program consumption was 13.5 million cans. The coefficient of correlation between the estimated and actual MRI distribution is a quite high .96.

Variation was also measured on a program by program basis. For 65 percent of the programs, the estimates were within 15 percent of the actual consumption. For 11 of the programs, the differences were greater than 15 percent. This range of error is comparable to the variation reported when estimating TV program household and demographic ratings prior to the start of the new broadcast season.

Scan America also measured the predictability of its Buyergraphic ratings. The technique chosen was to predict which shows will be in the top quarter of regularly scheduled prime time programs one and two quarters forward. The result indicated that Buyergraphics predict as well as demographics. One quarter ahead, they correctly predicted 82 percent of the time and for two quarters forward were 73 percent accurate.

Do Volumetrics Discriminate?

There is increasing evidence based on Scan America, Campbell Soup, Kraft, and other advertiser tests that utilizing product ratings does improve the effectiveness of TV advertising scheduling. For example, George Mahrlig, Director of Media Services for Campbell Soup Company, was able to increase target efficiency by 14 percent when selecting programs on the basis of Campbell V8 Juice Guide volume index. The "V8" index is developed by indexing each household's "V8" usage volume to the average household usage as reported by IRI Research. In this way, Campbell Soup was able to select programs to increase exposure against "V8" Sales Volume potential.

Similarly Bill Landgraf, manager of advertising services for Kraft's Retail Food Division, has devised a system to allocate TV announcements to those corporate programs with the best environment. All Kraft corporate programs are analyzed on the basis of selectivity indices. For example, MRI Research shows that Mr. Belvedere is 12 percent above average for Minute Whip but is 18 percent below average for Breyer's Ice Cream. Corporate brand allocation is based on assigning brand commercials to the programs with the highest indices. Mr. Landgraf has estimated that through the use of this technique, Kraft has an eight percent lift across the board.

In an effort to test this approach, MRI secured the help of Ted Bates Advertising. The agency furnished MRI with a listing of all prime time corporate programs purchased by Mars Inc in the fourth quarter 1985. Our goal was to determine whether brand schedules that were selected in a standard manner (using age and sex criteria) could be improved during the corporate allocation stage by using product user and volumetric research.

Out of Mars Inc.'s 31 corporate programs which had been primarily selected based on audience age and sex criteria, eight had been assigned by corporate management to Kal Kan Dog and Cat Food. The actual Kal Kan schedule delivery was compared to other alternatives at the same budget. The alternatives schedules were based on selecting out of the corporate list those programs that delivered the highest product consumption, brand users, or product users as measured by MRI in its syndicated reports.

Selecting programs on criteria of volumetrics, consumption, or users does discriminate. The actual schedule was improved by selecting programs out of the corporate list to maximize consumption. When this was done, Kal Kan was able to cover 18 percent more product volume and eight percent more brand users for the same budget than did the actual schedule assigned by corporate management based on age criteria only.

What this eight percent to 18 percent advantage means is that it is possible for advertisers to extend their media values approximately $80,000 to 180,000 per million dollars of budget. If the goal of media is to maximize coverage of sales potential, a direct match is superior to buying indirectly by using age and sex criteria only.

It is for this reason that consultants like A1 Achenbaum and corporate marketers like General Food's Kent Mitchell are quoted in the Harvard Business Review, "The old methods of valuing the media may no longer apply. Media cost per thousand packages sold, say, may be more useful than audience cost per thousand."


While the Age of Single Source has dawned, the Traditional Single Source services can be effectively used now to improve the yield of media schedules. There are four reasons given by practitioners for not using product consumption, volumetrics, or users in TV program selection. These have been reviewed. It is our belief that using volumetric measurements now can help advertisers discriminate among program choices because: . Micro computer make the use of volumetrics practical at the corporate allocation stage. . Consumption and user data are stable and reliable from report to report. . Consumption by program can be predicted with the same precision as viewer composition. . The use of volumetric measurements discriminates by program and schedule. Translated to dollars, this means that brands may be able to secure $80,000 to $180,000 extra media exposures per million dollars of budget.

The implications of these findings are clear. We are in the midst of a revolution in audience measurement and the ability to evaluate data via micro computers. As more and more personal computers are placed in media departments, we expect to see a trend away from planning and buying solely on age and sex rating points. More buys and plans will be based on direct match against prospects in terms of users or volumetrics. If we're interested in covering our sales potential, a direct match will maximize our advertising budgets whether we are planning or buying for network TV, cable, radio or print.

The system won't change overnight. But if the 60s were characterized by the use of research covering age and sex, the 70s by socio-economic demographics, the 80s by geo-demographics and lifestyles; perhaps the 90s will be the era when media plans and vehicle selection will be based on consumption and volumetrics. The Age of Single Source will then be a reality.
COPYRIGHT 1989 St. John's University, College of Business Administration
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kamin, Howard
Publication:Review of Business
Date:Jun 22, 1989
Previous Article:Tracking research: the state of the art.
Next Article:The telephone as computer.

Related Articles
Spending patterns of men and women.
Britney tot? Don't be a dummy.
Tragedy of girl, 5.
Social Security: the largest source of income for both women and men in retirement.
40-hour work proposal significantly raises mothers' employment standard.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters