I recently completed consulting work for two US companies. The first, which we will call Midwest Stores, is a big grocery store chain. A year ago it repositioned its retail brand based on the findings of a conjoint study.
Conjoint research is very revealing. It allows a company to find out what consumers most want from its operation. In this case, Midwest asked 2000 customers to rate the importance of price, products and service.
While all were important, Midwest's discerning customer base actually valued product range above price and service.
Midwest repositioned accordingly, with the slogan 'Every brand under the sun', and increased prices slightly to offset increased distribution and marketing costs. However, sales have shown no discernible increase.
The second company, which we will call Baxters Yoghurt, also had problems with sales. Baxters is a mid-sized dairy, which has lost more than 20% market share in the past three years and was unsure what to do about it.
Its ad agency conducted focus groups with existing customers, which revealed a growing sense of dissatisfaction. Everything from flavours, quality and product sizes to its latest ads were cited as reasons for unhappiness.
Baxters marketing director was unsure what strategic decisions to take to restore the brand's health. There were so many issues and, worse, the managing director was openly dismissive of the 'fuzzy' focus group results.
Both companies were struggling because their market research was deficient, but ironically each held the solution to the other's dilemma. Midwest is a classic example of a company that solely uses quantitative data.
This sort of data is weak because it is predicated on measuring a set of questions. This means it is only as good as the options offered to the consumers surveyed. While it may be possible to prove X is more valuable than Y, this result is flawed if the consumer actually wants A, B and C.
The company conducted its survey correctly, but tested a very limited set of alternatives provided by the marketing team. To truly understand customers it needed to postpone its measurement stage until it knew from customers, in an open-ended way, what they wanted.
In contrast, Baxters understood its customers' problems well, but by relying exclusively on qualitative data, it was unable to identify which were most prevalent or important to them. 'Qual' data is weak, as the sample sizes are never representative.This means that magnitudes cannot be developed and, quite rightly, senior management are often uncomfortable executing a strategy based on insights from a small and unrepresentative sample.
Any company that relies exclusively on either qual or quant data will eventually fail. The secret is to combine them. Use a qualitative method, such as ethnography or focus groups, to learn from your customers. Then turn these learnings into a questionnaire and conduct it across a representative sample. In the research game what is needed is neither fish nor fowl, but a combination. A kind of fishy-fowly stew is the way to understand the market.