Surveys can be used to assess opinions on our messaging and the propositions we put forth into the market. It makes sense that we would want to leverage statements that resonate with the thought process of our consumers and prospects. This is where consumer and B2B market research can inform the creative process, leading to more effective marketing efforts.
There are several methods for assessing a respondent’s feelings about what they feel is important to the purchase process and how various message alternatives might impact those feelings. Importance scaling is one of the more common techniques. An example survey question would be:
How important are the following factors in your decision to choose a technology provider?
Ease of use
Company reputation
Product/service selection
This list would address the most common factors. Each factor would be scaled ranging from ‘not important’ to ‘very important’. The difficulty with this process is respondents quite often view everything as mission critical. Statistical analysis requires variation in the data, and if all responses are clustered at one extreme then it can become impossible to assess what’s most critical.
The Max-Diff process (short for maximum difference scaling) gets around this by creating a series of experiments where respondents are shown the factors in small groups and are asked to select a most and least important factor. This is repeated several times thus allowing all factors (or messages or value propositions) to be tested. This is an optimal solution, however to deploy Max-Diff requires special software and knowledge of experimental design.
Ranking the statements or factors is another option. However, this method loses its effectiveness if the number of items to be ranked goes beyond 5 to 7. It also produces ordinal data, which limits the types of analysis available to the researcher.
Allocation is an option that is easy to deploy, understandable and produces metric data (a must for more advanced survey data analysis.) Respondents are asked to allocate a total of 100 points across the number of factors or messages. A respondent can allocate between 0 to 100 points to any given message or factor. The more points allocated the more favorable they are to the statement. Best practice is to limit the number of factors/statements to 5 to 7. If necessary a larger pool can be randomly allocated into smaller groups and the exercise can be repeated for each group.
Allocation problems generate metric data which can be used in a variety of statistical procedures. Your survey should capture at least 3 – 5 profiling variables which can be used to create comparisons for the allocation data.
Photo:"Unbalance Of Golden Scales" by Vichaya Kiatying-Angsulee (http://www.freedigitalphotos.net/)