As marketers it is our job to understand what resonates in the minds of our consumers. With any luck this insight can be carried over to our customer acquisition efforts as well thus ensuring continued growth. From a market research perspective there are numerous ways to open the door and peak inside the minds of our consumers and prospects. Normally, I prefer a simple path, but in order to maximize the value of the information obtained we often delve into our toolbox of methodologies.
MaxDiff is one of those methods that are gaining increasing traction in the market research world. The method was developed in 1987 by Professor Jordan Louviere. It affords researchers the ability to use experimental design techniques in assessing the factors respondents consider to be most or least important in a their purchase decision.
The advantage MaxDiff has over other ranking or rating techniques is that it prevents respondents from saying that “everything” is important. If scales are being used to measure attribute importance then the risk exists that participants will give all attributes a “seven” in terms of importance. On the other hand, if traditional rank order methods are used then in reality respondents can only give credence to a limited number of attributes before the validity of the data suffers.
MaxDiff presents a randomized sub-set of attributes to the respondent. They are then asked to select which attribute is considered to be most and least important. This is repeated anywhere between 10 and 20 times. After all is said and done the data that comes out of the process is metric. Here the researcher can make legitimate claims to the magnitude of one attribute’s importance relative to another in the context of a purchase decision, value proposition or advertising claim.
MaxDiff does require advanced programming capabilities and more time to create and implement than other methods. However, the increased precision of the data more often than not outweighs these additional costs.