1
$\begingroup$

I'm working on a segmentation of individuals who have particular purchase behaviors. The net result is a set of 20 segments of people based on their consumption levels across two axes. An example would be:

Axis 1: Light to Heavy consumers of yogurt Axis 2: Light to Heavy consumers of dairy

Each segment has their own sub behaviors and demographic breaks, etc.

Are there any helpful visualization techniques for thinking about bucketed data with many sub-breaks for any one question? Right now we are frequently using a table with 20 spaces, but it's frequently hard to work within.

If each group can be grouped into 4 demos that we want to show, are there any good data visualization examples that would map onto something like this?

$\endgroup$
1
$\begingroup$

Twenty sounds like a lot, and I'm guessing that you need to distill or group those into particular over-arching segments. Even Nielsen's PRIZM, with 66 segments, divides those into Social Groups and Lifestage Groups. It also sounds like your axes might be duplicating data - are people who are heavy consumers of yogurt by necessity also heavy consumers of dairy? Or is the point that there are lots of people who don't do dairy at all but will do yogurt? In other words - what is the story you are trying to tell (or trying to get the visualization to reveal more clearly)? It's a little tough to give solid recommendations without that, and if you just don't know, then maybe Excel spreadsheets are your best bet.

Apart from that - what measure(s) are you trying to display? The number of people in each group? The average household income of people in each group (e.g., the people who buy yogurt tend to have average household incomes 30% higher than the people who buy process cheese)? Different visualizations might be appropriate for each one. A bubble chart might help with this - imagine that the color of a bubble indicates "light" or "heavy" consumption, the size is indicative of the number of people in a group, and maybe the X or Y position indicates membership in a categorical segment. A heat map might also help.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.