This paper will tell the story of the smallgroup of people who, in France, have been in charge of the measurement and theappreciation of the audience of television, and had to invent audience research,to organize it and to communicate its results to "clients" whodepended on it much earlier than usually assumed: television managers andprofessionals, public authorities, and, last but not least, advertisers. The paper will explore both change andcontinuity. First, change: the professional origin and training of measurerschanged much over the years. In the early days, they could have an almostliterary profile. The first person in charge of the audience at the RadioTélévision Française was a teacher of philosophy. His followers had a formationin sociology and semiology. They all insisted on the fact that they were not"simply" measurers, and also worried about appreciation, quality,culture. They always figures did produce figures, but rarely only about thenumber of people present in front of the screen, mostly about satisfaction,appreciation, preferences for given genres, viewing habits. As there was onlyone channel – until 1964, with the number of TV sets rising sharply, ratings,in the modern sense, were not critical.Things started to change gradually. In 1974,the public broadcasting corporation was divided into several companies,including three competing channels. The service in charge of measuring audiencewas put under the direct authority of the Prime Minister. Audience figuresplayed a part in the distribution of resources, not only advertising but throughthe license fee sharing. However, the law also provided a clause about an indexof quality, which never functioned satisfactorily, although the service incharge of audiences put much effort into it. In 1985, around the time of deregulation,change came. From outside, this was translated into the rise of daily,detailed, fast produced figures of the audience through audimeters, thenpeoplemeters. Those figures become highly controversial; a popular book on TVwore the title: "The dictatorship of the audimat". But audiencemeasurers did not turn into dictators. They were, undoubteldy, more consideredas technicians. Their competence in statistics became crucial. But they werealso negotiatiors, consensus builders who have to work in an atmosphere ofgrowing suspicion as the revenues of television depended now mainly, if not only,on audience figures. However, continuity was there aswell. The need for effective mediations of the audience existed from the start.Those mediators, figures, reports, played several roles. Particularly, and thisis true until two days, they provided channels managers with a source of“para-democratic legitimacy”. For the “profession” of measurers, this meansthat they have always played an important roles, as spokespersons of theaudience, equipped with an almost magical kind of knowledge: the power to “read” the will and whims of a mysterious, anonymous mass of viewers.