Product-$ales Effect-Test

Which Image- and $ales Effects do Products or Product Changes achieve?


Product tests are widely perceived to belong to the easier tasks in market research. Often companies run their own product tests. The research design stays nearly always the same; questionnaires are short and do rarely changed.

73 percent of all new product launches end as flops and most relaunch attempts fail to achieve their intended objectives. Indications are that a lot of these product failures are caused systematically by mistakes in product tests. The Product $ales Effect Test has been designed to solve the following identified major problems:

  • Product tests are still often carried out as "blind tests" (without brand identification). Blind test results as basis for product launch decisions however can be misleading if brands have created spe-cific product expectations. The well known "New Coke" flop was caused by such a mistake. More than 150.000 people preferred in blind test a new, sweeter product formula to Coke's original recipe. However "New Coke" failed in the market because most drinkers expected and preferred the classical less sweet taste experience when choosing a Coke.
  • Often the duration of product trial is too short to take adequately care of the so called "novelty effect". This can result in euphoria if people try products with lower usage frequency or prefer more variety in food and beverages. An example is grape juice which usually is significantly preferred in studio tests to apple and orange juice, but subsequently rarely bought in the market.
  • In case of intended relaunches often the reaction of the brands' current user base is not analysed separately. If a larger part of loyal brand users perceives the new product formula less favourable it is usually better to maintain the current product. Because experience shows that dissatisfied current brand users disappear quickly, but it takes substantial time and effort to win new customers.
  • Often new product formulas are tested versus the market leader. But due to the ongoing market fragmentation market leaders usually have a user share of 30 percent or less. If you use representative consumer samples for product testing, for the majority of the respondents the comparison of a new product formula to the market leader to is irrelevant, because they have a different individual main brand. - For brand choice individual customers use a simple bench mark: the actual brand most often purchased, thus reducing complexity. Because the current main brand offers individually the "best problem solution" of all known alternatives. It determines which perceived combination of emotional benefits and factual features at what specific level satisfies optimally the individual need structure of all competitive offers specific respondents are aware of. The onus is on the "new" (or relaunch) brand to prove that it is better than current main brand. If it fails to establish this superiority in the perception of recipients, individual consumers will feel no motivation and see no reason to switch from their main brands on a long term basis which have served them well in the past.
  • Verbalisations alone are often insufficient to measure multisensory product experiences.
    Most of our senses are usually involved; their impressions often interact and can be influenced by a lot of other factors (eg. temperature).
    • Experience has proven that a lot of respondents use the terms "bitter" and "sour" not precisely enough. This is partly due to the fact that both taste components often appear in combination (e.g. in lemons, grapefruits).
    • Taste experiences and smell often interlink in product perception. Overall approximately 400.000 different smell particles do exist (Vanek 2004), of which humans can distinct about 10.000 (El-Sohemy 2007) and animals significantly more. But language has only a limited number of words to describe and measure different odours. The developed "verbal distance models" of similar and dissimilar smells have not perford well in practical applications (e.g. Amoore, 1964; Henning, 1924)
    • Experience has proven that respondents can verbalise mouth feel and acoustical impressions only inadequately to provide R&D personnel with sufficient insights and support. Also language has its limitations if temperature and pain impressions are described by respondents. A larger quantity of product tests does not measure these often behaviour relevant dimensions.
    • Visual impressions of the packaging, of the product in the packaging and of the product on its own often have significant influence on product experience and brand choice of prepacked goods. Mental anticipation can influence or even dominate the "objective" taste experience. Considerable research experience has proven that orange juices in darker colours are usually significantly better rated than lighter coloured ones – despite identical "objective" taste experience. - Beck's Gold in a clear (UV resistant) glas bottle was perceived to have a significantly milder taste than competitors such as Warsteiner or Krombacher, despite the fact that the bit-ter ingredients in these beers were quite similar. (Shaw, Schipke, Mayer de Groot 2004)

Therefore the Product-$ales Effect Test uses symbolic attribution in addition to verbalisations.

The results of conventional product tests usually do not provide information concerning future market reaction to a new or changed product formula and therefore can cause sub-optimal decisions. Even worse they can send wrong signals which may steer product development efforts in false directions for longer time periods. Therefore it is important to use product sales effect tests which also measure and predict image and behavioural effects reliably.

The following case studies indicate the opportunities:
Beck's Gold Case Study