Origins

     Arthur C. Nielsen (1897-1980) founded his marketing com­pany in 1923 after serving briefly as a naval officer in World War I and working with two Chicago companies. With six employees and $45,000 in capital from Nielsen's former frater­nity brothers, the company specialized in performance surveys of industrial equipment. The company went bankrupt twice in its early years. A decade later, Nielsen expanded his service by launching a continuous market research service, the Nielsen Drug Index, to chart the retail flow of specific products. The Nielsen Food Index soon followed. Both were based on the same premise (which was later applied successfully to broad­cast ratings): carefully develop a sample of stores and visit them periodically to measure unit sales through audits of pur­chase invoices and shelf stock. When projected regionally or nationally, these data provided a measure of sales that could be related directly to marketing efforts.

Nielsen entered the radio audience measurement business at the request of clients who found the food and drug indexes useful guides and desired the same assistance in purchasing radio advertising time. In 1936 Nielsen acquired the rights to a mechanical device developed by twoMassachusetts Institute of Technology professors, Robert Elder and Louis Woodruff. The "Audimeter" made a graphic record on a filmstrip, providing a continuous record of radio receiver use-when it was on and to which station(s) it was tuned-over a month-long period.These early meters were both costly and cumbersome, espe­cially as the tape had to be picked up by Nielsen personnel before the tabulation of results could begin. The tapes were then shipped to a Chicago plant where they were "read" by specially designed machines. After modifications to the meter, their use was subject to intense experimentation for four years in several Midwestern states.

Radio Ratings

     The Nielsen Radio Index (NRI) ratings service, based on the meter system, was introduced commercially in December 1942 in competition with the then dominant "Hooperatings," which used telephone surveys of sample homes. A key advantage of the Nielsen meter was that its sample (initially just 500 homes in the east-central portion of the U.S.) was not restricted to tele­phone-owning homes; this was important at a time when upward of a third of homes in some areas lacked the instrument. By 1946 the NRI had expanded service to some 1,100 homes over most of the country. NRI also introduced an improved meter with a mailable tape (it provided measures over two weeks) to speed delivery of the resultant ratings and to render personal staff visits to Nielsen sample homes unnec­essary. The streamlined process-which could measure four separate radio receivers-allowed expansion of meter-based ratings to both FM listening and television watching.

By early 1949, the NRI sample had expanded to cover virtu­ ally all of the country except for the Mountain time zone, which was especially expensive to serve. In early 1950 Nielsen purchased the Hooper national radio and fledgling television ratings services (Hooper continued local radio market ratings for several years). By this point, Nielsen's "methodology, finan­cial position, organization and widespread industry acceptance rendered him nearly invincible" (Beville, 1988). A year later, the NRI sample was up to 1,500 homes-and its charges to adver­tisers and broadcasters had nearly doubled. But these were national (network) ratings, not local-market measurements.

The Nielsen Station Index (NS!) debuted in 1954 to measure household use of both radio and television on a local­ market basis. This service was not audimeter based, but rather combined the use of traditional diaries (in which audience members recorded their listening time) with a "Recordimeter" device, which signaled with light flashes and a buzzer when listeners should make a diary entry and at the same time kept a rough measure of when the receiver was on. This crude meter helped to validate the diary information provided. And the diary could provide what no meter then could-demographic information on the gender and age of the person listening. In 1959 computers were first applied to Nielsen ratings processing and analysis. By the early 1960s, NSI was measuring radio listening in more than 200 markets. But its seeming market dominance would be short-lived, for, as Hugh Beville writes,

In 1962 Nielsen discontinued quarter-hour ratings because of declining radio listening levels and the rapidly increasing number of radio stations. This cost many client cancellations, which sparked the NSI decision to abandon radio. Not only was television seriously diminishing prime-time radio audiences, but the advent of automobile and portable receivers, plus many new independent stations, was rapidly changing basic radio listening patterns. In 1963 the local radio service was discontinued (Beville, 1988).

Contributing to the end of Nielsen radio ratings was a series of congressional hearings into the ratings for both radio and television. Nielsen became a central target in those hearings, in part because of methodological questions about some of the company's means of ratings data collection. Nielsen's system measured only home viewing, not portables. The out­-of-home audience, as it became known, grew with the movement to the suburbs and the use of the automobile. In response to the changing radio audience, Nielsen created an Audimeter to be installed in automobiles. However, when his clients were unwilling to support the investment costs needed to upgrade, he decided to quit the radio business. The Nielsen Company decided to focus its investment efforts on the greater returns potential from television ratings. The end of Nielsen's radio services was a key factor leading to development of RADAR national radio ratings.

Later Years

After leaving the radio ratings business, Nielsen continued to develop its national television network and local-market ratings services. It introduced overnight ("instant") television meters in major markets in the early 1960s and slowly expanded the process to other cities and network ratings. In 1987 Nielsen introduced its still-controversial "people meter," which could measure TV receiver tuning as well as who was tuning in. With Arbitron's departure from television ratings in the late 1990s, Nielsen became the only source of both national and local-market television ratings.

When the elder Nielsen retired in 1976, A.C. Nielsen, Jr., became the company's leader. In 1984 he sold the firm to Dun and Bradstreet, which in 1998 split the marketing and media research aspects of the company. The latter was sold to Lucent Technology and then spun off to a new corporate owner, Cognizant Corporation, in mid-1998. At the end of 1999, Nielsen Media Research was purchased by a Dutch company, VNU NV.