Advertising research at the turn of the 20th century was best exemplified by the theoretical research of Northwestern University professor Walter Dill Scott, who conducted controlled experiments using fake ads to assess consumer reactions to advertising. Daniel Starch, however, was interested in studying how consumers reacted to ads they encountered in real magazines under ordinary (not controlled) circumstances.
At the time, the idea of trying to measure readership was disparaged by many advertising practitioners, but Mr. Starch developed a magazine readership measurement method that for the first time applied recognition measures to advertising research. His findings were published in 1923 in "Principles of Advertising," a tremendously successful book, which garnered more income for Mr. Starch than did his job as a lecturer at the Harvard Business School.
Based on this success, Mr. Starch created Daniel Starch & Staff Research Co. in 1923.
"Principles of Advertising" was a landmark book in advertising, influencing the structure of ad textbooks for decades. Its popularity sprang from its comprehensive, pragmatic approach. The text addressed five fundamental questions advertisers face: To whom may the commodity be sold? By what appeals may it be sold? How may the appeals be presented most effectively? By what media may the appeals or advertisements be presented? What is a reasonable expenditure for advertising the commodity?
Mr. Starch soon became a leading name in the ad industry. In 1924, the American Association of Advertising Agencies asked him to be the director of its new research department. He agreed to work for the association two days a week, while still teaching at Harvard and working in his own company.
While research director at the Four A's, Mr. Starch moved into the study of media beyond magazine advertising. He studied magazine and newspaper circulation and undertook pioneering research in radio. In 1928, the National Broadcasting Co. commissioned Mr. Starch to undertake the first project to estimate the size of the national radio audience. Using probability sampling techniques, he estimated the number of households with radios to within four percentage points of the figure determined by the 1930 U.S. census.
Mr. Starch's research in print media provided some of the first estimates of duplicate readership, a measure of the number of people who may read more than one medium or specific media vehicle. At a macro level, duplicate readership is an estimate of the number of people who would read both newspapers and magazines, or listen to the radio and read magazines. At a more specific level, it estimates which magazines a person read and which radio formats he or she listened to.
Beyond the specific findings of his Four A's studies, Mr. Starch's greatest contributions to the field of advertising were his methodological innovations.
In 1932, Mr. Starch formally started collecting data for his ad files. At first, a small number of advertisers and magazines commissioned his company to study their readership. Top national magazines, including Cosmopolitan, Ladies' Home Journal, McCall's and The Saturday Evening Post subscribed to the company's Continuing Readership Program, and these magazines soon found that its data provided a wealth of information about how to attract magazine readers' attention to advertising.
Daniel Starch & Staff went on to employ hundreds of interviewers across the U.S., using the information gathered to measure readership for thousands of ads.
Although technology has brought about changes in the techniques used to collect data, the Starch methodology is basically the same as it was when Mr. Starch created it in 1923. The process begins with field interviewers identifying a statistical sample of people who have read a particular issue of a magazine. Mr. Starch tested samples with sizes ranging from less than 100 to 10,000, finding that that stability of results was obtained with sample sizes of approximately 200.
For a particular issue of a magazine, 100 to 200 men or women are interviewed. The gender of the respondents is determined by the gender of the target audience for the particular magazine; each sample is collected from 20 to 30 different regions of the U.S.
Once a respondent has been identified, the interviewer then turns the pages of the magazine, waiting for the respondent to indicate if he or she had read or seen a particular ad. To guard against boredom and order effects, each interview starts at a random page location within the magazine. If a respondent indicates that she or he has seen an ad, the interviewer probes the depth with which the respondent read it. For each ad of one-half page or more, the interviewer reports whether it has been "noted," "associated" (with the product or advertiser name) or "read most" (read more than half of the copy). Additional scores are also recorded for the major visuals and the ad signature.
As a benchmark for comparison, the company calculates Adnorms, which are the average noted, associated and read-most scores for the previous year for all ads. Adnorms are calculated for each product class, and within each product class they are measured separately for men and women for each magazine. Adnorms are also calculated separately for specific types of ads, such as half-page, page, b&w and color.
During the 1930s, rival researcher George Gallup began assessing print ad effectiveness using an aided recall measure. Mr. Gallup's aided recall approach gained popularity over Mr. Starch's recognition method, primarily due to its easy adaptation to broadcast media and the prevailing assumption among advertising analysts that recognition measures resulted in an over-reporting of ad readership.
Even as recall gained in popularity, Daniel Starch & Staff was busier than ever, for magazines still found value in its Continuing Readership Study.
Mr. Starch died on Feb. 5, 1979, at the age of 95. The company he established has evolved through several mergers with other research companies to continue the research legacy established by its founder. It acquired C.E. Hooper Co. in 1969 and expanded its international presence with the acquisition of International Research Associates in 1974. In the early 1980s, the company, then Starch/INRA/Hooper, acquired the public-opinion research firm Roper Organization, which continued as a subsidiary of Starch under its own name until 1993, when the company name was changed to Roper Starch Worldwide.
In 2001, Roper was bought by NOP World, a market research division of U.K.-based United business Media, and merged with its Audits & Surveys Worldwide to become Roper ASW.