34 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS aim is to sell a passion fruit drink, you would examine the reactions of the small section of the population who do like it. It thus depends on the objectives of the test and the interpretation of the results, but these results must be available to everyone concerned, so that they may be aware of the reasons both for and against, and what the test objectives were in precise terms. DR. M. N. RODaER: How important do you consider the actual final preference score in product tests? TI•E L•CTUR•R: One must define the objectives very precisely at the outset. If you have basic market in[ormation, you will know what kind of people the market is composed of, where they are, what and how often they buy, how they use the product, etc. Your survey will also provide a picture of brand usage and how this differs from brand to brand you will also have the brand images of the main brands in the market. If, in the light of this information, you have to develop a product to compete directly with the market leader, then you would test against the market leader. But if there was a gap in the market for a product with certain qualities which the market leader lacks and no other brand appears to provide and you make a product to fill that gap, you would test it amongst those people a product of that kind might satisfy. They might not be easy to locate, but this would be your objec- tive. You might still decide to test your product against the market leader, and although you would not then expect it to be preferred overall, you would wish to ascertain its performance with those people you think might like that kind of product. In this case if the product did not beat the market leader, you should not necessarily scrap it. One must interpret the results in the light of the objectives. Mi•. R. CnARK: How in fact do you identify these minority or target groups in the context of consumer evaluation? TI•I• L•CTUR•R: Usually by interview, but it depends on your aim. If one is seeking people with a particular personal characteristic they may prove difficult to isolate. It may be necessary to undertake the tests in two stages, first a test using a representative sample of the population to pinpoint the people who like your product and dislike the market leader, and second to find out what kind of people they are. MR. N. F. E. BLACI(•O•E: Speaking as a product development chemist, I am concerned about the comparative insensitivity of the typical user panel. With the competitive situation that exists at the moment in the toiletries and cosmetic fields it has become increasingly difficult to make a significant improvement to our existing product types. Such improvements when made, however, can be demonstrated by laboratory test methods and supported by tests on expert panels. When it comes to trying out the improvements on a larger consumer panel the difference is reduced virtually to nil due to the inability of the panel members to discriminate. I feel that in such a case, should the improvement (enhanced holding power in a hairspray, for instance) be pointed out to the user - as would be done by advertising if the product were marketed - then the panel might be more able to notice the improvement. Can anything be done to modify consumer panels along these lines? •I'•E L•c'ruR•R: Frankly, I do not think •ve xvant to. In developing this product, you are very close to it you think it is important. The average consumer who is using a large number of products every day and bombarded with advertisements for many more, does not accord it the s•m• imp)r•nz•'. Th ' attribute xvhich you think vital

PRODUCT TESTING IN THE SPHERE OF PRODUCT DEVELOPMENT may not be one to which the average consumer attaches any great weight, and the product test results may reflect this. In your example you quote a high percentage of no preference. This is a useful indication whether or not the public are concerned about the differences between the products. I would not interpret it to mean that there is a big difference between them that they had not noticed, providing of course that the test had been properly designed and executed. MR. N.J. VAN ABBg: In page 25 (b) you refer to re-interviewing a minority to get further information. It seems to me that one of the bugbears of all product testing is memory and its fallibility. Surely you are more than doubling the variants when you go back again after the interval needed to analyse the first results. Do you think this is ever really justifiable? TI•E LECTURER: This is a valid point, but if you want to find out what these people think one must go back to them, or repeat the test asking the appropriate questions of those who expressed a certain view. I entirely agree that memory can well bias results and one must guard against this in the test design as far as practic- able. If this eventuality can be foreseen the appropriate questions can be built into the recall interview. MR. R. CHUDZII/OWSI/I: What would be the best approach to panel testing of products, the concepts of which involves complete re-education of the public, for instance new modes of application? THE LECTURER: YOU may have to educate them how to use this product by demonstration, or possibly by using an explanatory leaflet. MR. R. E. ECIrTON: To what extent do you find that consumers are able to relate perfumes to product claims? TI•E LECTURER: This is one of these difficult areas where more work needs to be done, and this applies not only to perfumes but to taste, and the feel of things as well. We lack an adequate vocabulary to convey to people exactly what xve want them to understand they in turn are unable to properly communicate their impressions to us. A great deal of experimental work has already been undertaken in this and allied fields but more research is needed. Most people are familiar with basic flavours and able to distinguish betxveen sweet and sour, and so on. As yet we have not, however, created a satisfactory vocabulary for perfumes. Mt•. R. E. EcKTo•: Do you find that a preference for a particular perfume changes, or do people tend to choose a perfume on their liking for it, irrespective of the detailed claim for the product? TI•E LECTURER: I think that people are aware of broad types of perfume which they associate with particular uses both for shampoos and for other products. For instance, a floral perfume would not normally be associated with a medicated sham- poo.