Some advertisements are designed to produce direct results, such as sales or inquiries. In such cases, a basis exists for measuring accurately the advertisement’s worth.
Inquiry tests may be handled in several ways.
1. The same offer may be placed in different pieces of copy placed in different issues of the same medium. The offers are so keyed that they can be traced to the specific advertising copy. Assuming that all other factors remain constant between issues – a difficult assumption to make the difference in the number of inquiries received should indicate the pulling power of the different pieces of copy.
2. The same offer is placed in different advertising copy that appears in different magazines or newspapers. The assumption is that differences between media are either negligible or can be corrected for (say, on the basis of circulation), and that the adjusted returns will indicate the best copy.
3. The same offer is placed in a medium that provides a split-run service. This is a procedure whereby half the copies of the magazine or newspaper contain one piece of copy and half contain another piece of copy. This is accomplished in a systematic way so that the two pieces of copy reach similar audiences.
It must be remembered that a large number of inquiries does not mean that the advertising is successful unless inquiries are the sole advertising objective. If the advertising objective is to help the long term buildup of a general attitude, the number of inquiries may not necessarily be a sound basis for judging the advertisement’s effectiveness.
Sales Test: Such tests can be conducted using a variety of experimental designs. They are sometimes done as part of a test marketing program, in which case the number of cities is typically small. Despite their appeal such tests have important limitations: they measure only short term effects; the measuring instruments may not be sufficiently sensitive to detect relatively small differences in sales; they are time consuming; it is hard to hold all other factors constant and they are quite expensive.
Sales tests may be simulated in a variety of ways – for example consumers are exposed to alternative pieces of copy through point-of-purchase displays or direct mail. In the case of point-of-purchase displays, the need to test two advertisements would require two matched store samples. Product sales would be measured in each store before and after the introduction of the appropriate display. The changes in sales between the two periods for the two store samples would be compared and that group with the largest increase would be presumed to have the best copy. Similar experimental design studies can be conducted by mail, using coupons.
A some what similar test involves the use of two matched groups, the first of which is shown a group of TV commercials in a trailer located next to a supermarket. The second group does not see the commercials. Both groups are given tickets with a redemption value that group members use in the supermarket to buy whatever brands they wish in the product category involved in the test. The store, through its check out electronics, records the purchase made by each group, thereby providing the data needed to measure the effects of the test commercials.
Behavior Scan, a service operated by Information Resources, Inc., measures the effectiveness of television advertising in a similar fashion. The company has installed UPC scanners in all supermarkets in its mini-market (test) areas. Members of a panel of households in each of these market areas have their purchases recorded electronically by presenting their ID card to the cashier. The panels are representative of the total market and serve as the basis for studying trial and repeat buying, purchase rates, and brand loyalty.
Using proprietary technology, test commercials are substituted for control ads at the individual household level within the normal viewing environment. Samples are selected, based on their purchase behavior history. Since alternatives advertising copy treatments can be delivered to different samples at the same time, it is possible to measures the sales effects of one treatment versus another. The problem with such a service is that it is difficult and expensive to use in the early stages of copy design.