Purposes in test marketing
Category: Marketing
Pilot launching is a readily understandable objective. Any pilot operation aims to determine whether or not the system works as well as intended and at discovering design errors and weak components. For example, the pilot launch may uncover production problems. Until now only prototypes have been made probably on experimental plant. The pilot requires higher volumes and this will test the system. Occasionally, severe problems will be revealed as in the case of the plastics company that developed a revolutionary honeycombed product with tremendous impact resistance.
| Pilot launching | Marketing mix testing | Sales testing |
| Logistics check: | Marketing mix check: | Sales check: |
| investigation of | investigation of mix | generating data on |
| production, product | components and | penetration and repeat |
| packaging and | interrelationships | |
| distribution problems |
Their prototypes received much praise and many diverse applications were discovered in transportation and packaging. They made the prototypes by pouring the substance into small tanks containing metal rods and allowing it to set. For large-scale production however they grossed up the size of the tanks only to find they could not extract the rods without breaking the honeycomb. Another example is the Food Company that was encouraged at the product testing stage because respondents found the distinctively shaped glass container impressive. Unfortunately, the preparations for the pilot launch revealed that the glass manufacturer could not produce to the desired specification in the requisite numbers.
Another illustration of a pilot test revealing production difficulties is provided by Cadbury’s Wispa chocolate bar. This was put on test in the Tyne Tees television region in late 1981, but was withdrawn in a matter of weeks because the production plant could not cope with unexpectedly heavy demand. A decision was taken to invest Ј12 million in new plant and the product was launched nationally in 1984 (Huxley. J 1984).
Consumer reaction to the product can also be more realistically assessed. Product testing, even in natural settings, is artificial. The pilot launch gives the opportunity to derive consumer ratings and to study how they use it. This may reveal that they do not use it in the predicted manner or on the predicted occasions. Instructions for use on the package may be shown to be inadequate; some consumers may be using the product inappropriately. Household cleaning products may be applied to some materials with deleterious effects, and foodstuff wrongly prepared or incorrectly stored. Of course, not all this feedback need be negative. Unforeseen uses in non target segments could appear and product improvement result.
Pilot testing will also allow the distribution system to be tested. Storage and handling characteristics of the products may be of interest. Silly problems may be encountered such as an insufficiently strong glue being used on the outer cases. Deeper problems may be uncovered in, say, a new palletised handling system.
Retailers’ attitudes and behaviour decide the ultimate fate of any new brands The test market is a forum for receiving information on these. Their willingness to stock is the acid test for a brand’s launch. Obtaining ‘listings’ in the key multiple accounts is crucial.
Marketing mix testing is an experimental approach to evaluating the final marketing mix for the new product. As Corkindale and Kennedy illustrate however, there are frequently two major failings in the conducting of such marketing mix test marketing:
The purity of experimental design is often lacking. The sample sizes used are often inadequate.
Corkindale and Kennedy illustrate for example that to be 95% sure of detecting a 1% change in market share one would need a sample of 6000 product users. If product users were only 10% of the sample then the sample would have to be ten times larger. For many products the users could well be less than 10% of the population making the multiple much higher. Few people, other than government, would ever conceive of using a test sample size of 60 000, few even go to 6,000.
In planning full test marketing exercises for whatever specific purpose, four important factors should be borne in mind:
(i) Test areas should be representative (i.e. they should be as nearly as possiblea microcosm of the total population).
(ii) Advertising for a test market should be the same as planned advertising for anational launch.
(iii) Salesforce activity in the test market should be typical of planned activity for anational launch.
(iv) The test product should not receive an inordinate amount of managementtime. Sometimes a special management and sales taskforce is developed toimprove retailer and middlemen co-operation. Again this needs to mirrornational plans post launch.
In ensuring that test areas are representative, demographic and socio-economics profiles of possible test areas are compared to national profiles. Product class history and consumption trends are also compared. Some manufacturers choose their test areas on different criteria. Confectionery companies often use central Scotland as a test area because it has the highest per capita consumption of confectionery in the UK. Presumably they feel that if a new brand is successful there it will be so elsewhere.
In making advertising effort in the test correlate with post launch advertising plans, an important aspect is the level of advertising expenditure. A simple method would be to scale down the spending on a population basis. A more complicated approach would be to scale down according to the total exposures expected from advertising in different media. That is to have an idea of the number of times the campaign would be exposed to individuals at the national level say 10 or 20 million exposures. Then the estimate would be made of the segment size in the test area and the expected proportion of the national market that that segment represents. Applying the national exposure levels to the individuals in the tested area would yield the desired total exposures there. Media would then be checked to ascertain the cost of delivering those exposures in the test area and this would provide the scale of the advertising expenditure.
A second aspect of test advertising is the choice of media. Town tests cannot employ television in the UK. National press is also difficult to scale down although some national magazines do offer regional splits. The larger women’s magazines do have a reasonably controlled distribution which might allow area testing. The local and regional press are also anxious to demonstrate their usefulness in test marketing, and offer special deals to encourage this.
Ensuring that salesforce activity is typical can be problematical because the sales representatives are bound to be aware that management attention is focused on the new product. A disproportionate amount of selling effort could be devoted to the test item.