Almost all media models basically seek to maximize some measure of advertising exposure which is assumed to impact on sales. In brief, such models report, for a given advertising schedule, the number of individuals or households exposed and the number of times they are exposed. Since different schedules have different costs, the exposure output can be divided by the costs involved to obtain an effectiveness ratio.
In recent years there has been considerable interest in trying to obtain measures of the complete frequency distribution of exposure, the percent of the audience exposed once, twice, and so on. Such distributions are useful in measuring the value of one media schedule versus that of another.
But even assuming one can predict exposure with reasonable accuracy, not all individuals or households who are exposed are of equal value to advertiser. Some may be heavy buyers of the product class in question and, thus constitute prime prospects, while others may not be prospects at all. Further, how should successive exposures be weighed? Is the second exposure worth more than the third exposure? And if so how much more? How much time should elapse between exposures? How much forgetting takes a place over time, and what is the variation in forgetting between different exposures? And, of course, there is the differential effect between media vehicles as well as those related to length of commercial. While most models seek to address the above problems, they are handicapped by a lack of precise input data, especially with regards to current data pertaining to duplication, forgetting and accumulation.
Some models provide reach (extent of audience) and frequency of exposure for each media schedule. Others stipulate that advertisers state their objectives in terms of reach and frequency by target segments. In the case of the latter, the model output consists of a recommended media schedule by media vehicle, by the time period, and by number of units purchased, as well as information about the composition and size of the audience reached,, the total number of impressions, the distribution of exposure frequencies, and the cost.
Many advertisers obtain media exposure data that can be tied directly to brand usage. Such data clearly facilitate the selection of a media schedule, since a target audience can be selected and its exposure to various media vehicles can be determined as well. It is far better to match media vehicles directly with target market members (prospective or actual product or brand users) than to match indirectly on the basis of mediating variables, such as demographic characteristics of vehicle and target audiences.
In spite of the apparent advantages of media models, they are handicapped by a lack of precise input data, especially with regard to current data pertaining to exposure effects, communications effects, forgetting and response functions. In recent years, more emphasis has been placed on the adaptive interactive type of media model. Basically such models are designed to look at media problems from a management point of view. They are only as complex as the data and experience of the media planner. If for instance, the planner believes that advertising response does not vary by media option, or if the planner feels there are inadequate data on that point, then he or she is not forced to include this as part of the model. As management becomes more experienced, the model’s complexity reflects this fact.
People Meters: Since both passive meters and diaries have audience measurement problems, it is not surprising that considerable effort has been made over the years to find a better system. The most recent technological breakthrough is called a People Meter (an import from Europe) a remote controlled box with buttons that send signals to a small control box on top of a TV set. Each household member pushes his/her assigned button every time (s)he starts or finishes watching TV. The information about who is watching is recorded electronically and relayed to a computer that has on file the age and sex of each household member. Audience data at the individual level can be provided to clients within 24 hours.