Search
  • Kevan Oswald

Multidimensional Scaling


Multidimensional Scaling (MDS) is the process of representing consumer preferences and perceptions visually on a spatial map. It is associated primarily with brand positioning.


Some of common applications of Multidimensional Scaling include:

  • Understanding the position of brands in the marketplace relative to groups of homogeneous consumers.

  • Identifying new products by looking for white space opportunities or gaps.

  • Gauging the effectiveness of advertising by identifying the brands position before and after a campaign.

  • Assessing the attitudes and perceptions of consumers.

  • Determine what attributes the brand owns and what attributes competitors own.

There are essentially two approaches to Multidimensional Scaling with regard to measuring perceptions, similarity judgments and attribute ratings.


In similarity judgments, data is obtained from a survey where consumers are asked to compare how similar or dissimilar one brand is to another. For example, the survey may ask respondents to rate different brands of toothpaste.

With attribute ratings consumers are asked to rate brands with regard to how well the approach specific attributes.

Brands are then visually displayed using a perceptual map. At a minimum a perceptual map should contain eight brands, although perceptual mapping can be done with less, a well-defined map should have a sufficient number of brands to accomplish its objective. On the other hand a perceptual map with more than 20 brands is likely to be too cumbersome and require too many questions on the survey.


Below is an example of a perceptual map for several different car brands.


Perceptual maps with two dimensions--such as the example to the left--are the most common and easiest to interpret. However, perceptual maps with multiple dimensions representing multiple attributes can provide deeper insight.

In determining the number of dimensions, a stress measure is used. The stress measure determines the lack-of-fit; higher values of stress indicate a poorer fit. One of the most common stress measures is and elbow plot. The point at which an elbow or sharp bend occurs indicates the appropriate number of dimensions. Including additional dimensions beyond this point is unlikely to improve the fit.

In the perceptual map to the left we see that Motrin and Tylenol are perceived as being similar to each other, gentle, and good for children. From a consumer perspective they would be considered competing brands. Bayer on the other hand has less direct competition. Additionally, the length of each attribute line indicates the relative strength of that attribute. In this case, the two strongest attributes that distinguish pain relievers are “gentle” and “effectiveness.” The attribute “good for children” has a much shorter line indicating that consumers have a more difficult time distinguishing between brands on these attributes.

10 views0 comments

Recent Posts

See All