Commercials (on radio or television) may be tested at any stage in the process of developing a campaign. For example, concept testing is conducted during the planning and writing stages. The campaign's appeal or its basic assumptions may be studied in focus group discussions. Focus groups are groups of survey participants who are chosen for their relevance to the ​​research topic and guided through group discussions of that topic; for example, an advertiser wishing to test the potential effectiveness of a campaign to promote diapers would probably choose young mothers for participants in its focus group discussion.

     Each version of an ad to be studied in commercial testing is referred to as an execution. If more than one execution has been created for a campaign, the object of commercial testing is to determine which execution will be more effective in producing the desired results for the advertiser. However, only relatively large advertisers produce more than one execution for a campaign. In terms of the number of commercials submitted to testing, the most common situation is a test of a single execution. In this case, the aim of commercial testing is to determine how well the commercial performs with each of its potential target audiences. In addition, testing may suggest the kind of media purchase justified by the effectiveness of the commercial. A poor commercial may not justify heavy spending on media.

     What is measured in commercial testing? The most popular measures are called scorecard measures. They include recall, copy point recall, affinity toward brand or toward product and/or service, intent to purchase, and comparative brand preference.

     Advertising strategists assume that recall (remembering) is produced by attention, so measurement of a subject's recall of advertisements is actually a measure of his/her attention to those ads at the time they were presented. Decades of research into advertising indicates that, by itself, consumer recall of a brand or product name is not a powerful inducement to purchase. Because of the relatively simple process for measuring recall and the straightforward analysis of data collected, however, the measure continues to be popular. A typical study to assess recall involves recruiting (by telephone) a sample of adults who watch television during known hours and who are interested in the kind of radio station portrayed in the TV commercial purchased to promote the radio station. Each recruited respondent is sent a videocassette containing a television program in which the test commercial and others are embedded. The morning after viewing the video, an inter­ viewer calls to ask each study participant which products and services appeared in the video and what companies were represented. If 80 percent of respondents recall station KATT, then the recall score is 80 for KATT.

     When using copy point recall, interviewers ask specific questions about features of the targeted product or features of the commercial. If 60 percent of all respondents can recall the key points from the commercial, the copy point recall is 60.

     Affinity is a measure of what a person likes. It is assumed that when a consumer likes (has an affinity for) a product or sponsor, then he/she is more likely to purchase the product. Advertising research confirms a positive correlation between liking and purchase, but the relationship is not strong. Researchers often suspect that liking comes from previous exposures to the product. If a listener tunes to only one call-in show host, even if that consumer rarely listens to that host, the listener may report liking the host out of proportion to the amount of actual listening that takes place. Affinity can also be measured by scales that reflect several dimensions of liking. There is a growing preference for this kind of measure, as it may explain what considerations affect the magnitude of affinity in general. In the case of a call-in show host, listeners may like the fairness of the host and his/her treatment of callers but dislike the topics chosen for discussion on the show. This dislike for topics is likely to account for a lower-than-expected general affinity. Also, liking a product is often quite different from liking its manufacturer or dealer. The owner of a particular brand of car may dislike the quality of service provided by the dealer, so on the general affinity measure, the consumer reflects dislike although he/she retains a strong affinity for the car brand.

     The intent to purchase question asked of a respondent may be as simple as, "Are you more likely to listen to station KATT after hearing this promotion?" Or the question may be embedded in a scenario such as "Suppose that you go to the store because you have run out of milk. You are making a special trip just for this product. When you get to the store, your favorite brand is sold out. What are the odds that you will go to another store rather than try the brand in this commercial, which is available in your store?" Advertising research shows that intent to purchase is a complicated mental process for the consumer. If a young man has been wearing the same brand of jeans for a number of years and has been pleased with that brand, a long period of time will be required to effect a change of preference to another brand (assuming the brands are similar). So exposure to a test of a radio commercial for the new brand may produce very little change in the intent measure, but that small change may be significant because the consumer previously had never considered a change in brands. This is an especially important point when the products are radio stations, as a consumer's attachment to a radio station is rarely a rational process subject to logical argument. Transfer of emotional allegiances to radio stations may initially occur very slowly, then accelerate at surprising speed.

     The commercial test measure of comparative brand preference has considerable face validity (that is, it appears to be quite useful and reliable) for advertisers. But it is sometimes complicated to incorporate into a commercial test, and advertising research firms have risen or fallen in the past based on their handling of this measure. A typical comparative preference item might be, "If the brand you currently use was priced at $1.00 and you considered that price fair, what price would be fair for the product you heard about in the commercial?"

     Tests of commercials can be performed by nearly all local market research firms. They can also be contracted for by large national firms known for commercial testing, such as Gallup and Robinson or Mapes and Ross.