1. What is the problem in this case? What do you feel caused this problem?
The major problem was a non-sampling error and occurred somewhere around step 11 in the market research process - prepare and present the report. The information gathered did not appear to be clearly defined and they did not verify that they had a firm grasp on the information revealed.
* There should have been a report that indicated that the Classic Independent clothes buyer would NOT perceive the casual image of Levi's positively when purchasing suits.
* Secondly, they should have outlined very clearly the appropriate marketing mix for this type of buyer - e.g. higher priced, traditional style, independent outlets etc.
Had they clearly outlined this information, the marketing executives would not have been able to gloss over the importance of these factors and push this line under the Levi's brand nor distribute it through the department stores.
2. Could paying closer attention to the research have prevented the problem from occurring?
Yes - it was there but if they clearly outlined this information, the marketing executives would not have been able to "logically" over look this and follow the course of action that they did.
3. Do you feel the research was well conducted?
It does not seem that they successfully defined the market research problem. In defining the problem, I would have challenged the marketing executives assumptions - "the glue". Had the researcher questioned the executives about their beliefs regarding certain consequence of their proposed actions, they may not have been as certain e.g. distribution through department stores. There needed to be more "what if" type of questions. I would have specified constructs around their clients, who they were (Type 2) and link this construct to other constructs such as "comprehension of product benefits" or " attitudes or feelings towards brand".
Without more specifics around the survey and sampling methods, it cannot be easily determined if this was person/self/computer administered survey. This could affect how random the sample would be. In a mail survey, respondents decide whether or not to return the questionnaires - low response rates and potential biases if the respondents who decide to answer differ in some systematic way from those who do not respond. It seems likely that judgement sampling1 was applied to the original survey to get a sample which in some way represents the marketplace by choosing respondents who they felt were uniquely qualified to participate in the discussion groups.
4. What, if anything, would you have changed about this research?
By studying the executive's circumstances and resources, evaluating the competitive environment, factoring in consumer behavior considerations I would have developed a custom model of possible problem causes, solutions, and consequences. Further, I would have prepared an itemized listing of the information objectives agreed upon by the executives as essential for the purpose of resolving the current problems.
The response format to the original survey could have been scaled as a scaled response question is more complex and involves asking the participant to rate how strongly they feel about an issue on a scale (numerical or labeled). The brand image of Levis is a subjective property, which is very challenging for a researcher to measure. An ordinal scale or the Semantic Differential scale2 may be appropriate to indicate the gap between the consumer's opinion of the Levi's brand and what they were trying to present with the suits.
Also, Iwould have designed an inversely proportionate stratified random sample. Sometimes the population is somewhat heterogeneous which increases the likelihood that a number of individuals from all walks of like won't be included in a simple random sample. To over come this, I would have divided the sampling frame into groups of individuals, each of which is fairly homogenous and then drawn a random sample from each group. These groups are strata levels hence the name stratified random sampling. I would have intentionally weighted the classic independent strata more heavily. Care should be taken though when calculating statistics from stratified samples as they are weighted by the proportion of the sample in each stratum.
Further, I would have simulated test market if I have short timeline or a standard test market to test my product and its mix if I had the resources to make this happen e.g. a year and the cash flow.
5. Did Levi's follow the recommendations of the research? Why? Why not?
Creating a report on focus group sessions can be difficult as so much interpretation is left up to the discretion of the researcher. In this case, the interpretation of the focus group work not feeling comfortable with saying Levi's made their suit was considered subjective. The marketing executives surmised in error that they could be over come through a cute tag line like such as "Oh, and they can also make a good suit when they put their mind to it".
They not only ignored the findings of a predominant association of their brand with a casual image that perhaps could not be overcome, they also ignored the preferred distribution method of the classic independent buyer.
6. In your opinion what was the marketing mix for the new product when it was launched? Does this make sense given the research results?
* Price - higher priced (10% above competition)
* Product - traditional style, pinstripes etc.
* Promotion - Didn't indicate but should have been a more sophisticated magazines and models etc.
* Distribution - Chain store distributions (Should have been through independent outlets, while shopping alone.)
They were in error in their promotion and distribution strategies as they went against the shopping preferences of their target market. The distribution error was obvious but the promotion error was that there was not promotion strategy outlined. There should have been promotion efforts to this market e.g. appropriate commercials, print ads, point of purchase etc.
7. What other alternative strategies do you feel should have been considered? Does the research support your recommendation?
Another option to the discussion groups could have been depth interviews as they may have been able to flesh out the depth of association with the buyers perception and likeliness to associate Levi's with more sophisticated clothing such as suits.
Launching this product under a different brand would be a better way to go as Dockers has seen success under it's own brand.
1 Judgement sampling involves any sampling frame where the researcher chooses respondents based on some criterion other than randomness.
2 The Semantic Differential Scale is a specialized scaled-response question format that has sprung directly from the problem of translating a person's qualitative judgments into quantitative estimates.