Marketing Communications

POST-TESTING OF ADVERTISING 301

Post-testing of advertising

A post-test is a test of the eff ectiveness of a single ad aft er placement in the media. Post-tests

are only meaningful if there is a before measurement or a control measurement as a bench-

mark. Th ree types of post-tests can be distinguished ( Figure 9.5 ): measurement of exposure,

communications eff ect tests and measurement of behaviour.

First of all, the extent to which an ad has reached its audience can be measured. Net reach,

GRP, OTS and other exposure measures can be calculated. Normally, this type of eff ective-

ness measurement is done not only aft er media placement, but also as a part of the media

planning eff ort. Similarly, the amount of publicity generated in the press or sponsorship

exposure can also be measured and expressed in terms of reach or GRP.

In measuring the communications or message processing eff ects of an ad, two types of

tests are used: recall and recognition tests. A recognition test is a very obvious eff ectiveness

test. A sample of ads is presented to a consumer, who is asked to indicate whether he or she

recognises the ad or not. Th e underlying assumption is that ads can only be eff ective when

they are at least noted. A well-known recognition test procedure for print ads is the Starch

test. Some 75 000 ads in 1000 issues of magazines and newspapers are assessed each year,

using 100 000 personal interviews.^11 Consumers who say they have read a specifi c issue in a

magazine are interviewed at home. Th e magazine is opened at a random page, and for each

ad a number of questions are asked. Th e procedure leads to four percentage scores for each ad:

z Non-readers : the percentage of people who do not remember having seen the ad.

z Noted : the percentage of readers who claim to have seen the ad.

z Seen/associated : the percentage of readers who claim to have read the product and brand name.

z Read most : the percentage of readers who claim to have read at least half the ad.

Obviously, the Starch test is very susceptible to the test subject’s honesty. However,

research reveals that high ‘noted’ scores are positively correlated with a positive attitude

towards the brand ( r = 0.43) and a positive intention to buy ( r = 0.52).^12

In a masked identifi cation test , part of a print ad, usually the brand name, is covered. Th e

subject is asked if he or she recognises the ad, and if he or she knows what brand it is for.

Recognition and correct attribution scores can then be calculated. Obviously, brand con-

fusion can be measured too. Th e combination of recognition and correct attribution scores

leads to the useful score : the percentage of the consumer sample that both recognised the ad

and attributed it correctly to the brand advertised. In Figure 9.6 , an example of the results of

Figure 9.5 Post-testing techniques

M09_PELS3221_05_SE_C09.indd 301M09_PELS3221_05_SE_C09.indd 301 6/6/13 9:21 AM6/6/13 9:21 AM