Questions? (877) 643-7244
FutureNow Post
Thursday, Aug. 30, 2007 at 8:00 am

PRIZM Clusters Not as Predictive as Behavior

By Robert Gorell
August 30th, 2007

Don't be fooled by the PRIZMOn his Marketing Productivity Blog, Jim Novo drills down into the misconception that PRIZM (define) clusters can be more predictive than actual human behavior. A rare breed of marketer who understands both the statistical and behavior ends of the spectrum, Jim explains the danger of distorting one’s vision by gazing only through the PRIZM:

[...] what is the likelihood these households reflect the overall “label” of the PRIZM cluster? Combine this with the fact that for customer analysis, demographics are generally descriptive or suggestive but not nearly as predictive as behavior and you have a bit of a mess.

Here’s a test for you. It only requires rough knowledge of your neighbors, so should not be very difficult (for most people!)

1. What is your “demographic”?
2. If you were to walk around the block and knock on doors, how many households would you find that are “in your demographic”?

Right. Maybe a handful, unless you live in a brand new housing development or other special situation. Now think about walking your zip code, or walking out 10 blocks or so from your house in any direction, and knocking on doors. Do you find most of these people are in the same demographic as you are? Did you ever find the “cluster average” neighbor? [Keep reading "PRIZM Clusters Not as Predictive as Behavior"... ]

For a bit of background context, check out “The Marketing/Technology Interface,” where Jim discusses how the scientists among us are making it difficult for the
marketing folks to learn about using data by obscuring the actionable with the scientific. He pleas for “the technical side to be a little ‘less scientific‘ and the marketing side to be a little ‘more specific,’ so they can meet in the middle.

With all the talk of “engagement” as the elusive metric, chased by marketers and statisticians alike, it’s worth noting that Jim Novo & Bryan Eisenberg formulated it some 7 years ago in the Marketer’s Common Sense Guide to E-Metrics. It’s just that back in them days, it didn’t have such an elegant name: “Scanning Visitor Share/Index/Volume” and “Committed Visitor Share/Index/Volume“.

I reckon “engagement” has a better ring to it.

Add Your Comments

Comments (1)

  1. Yes, Jim’s post is really important. PRIZM clusters can be very tempting, and they are of course relevant in several situations. They don’t seem to be, though, when it comes to the Web. I have found them not to be predictive of behavior, patterns of response being the same across the various clusters (email program and loyalty program cases).

    “Engagement” is definitely the metric du jour. Some time ago, during his Visual Sciences days, Eric Peterson did some very interesting work about it, but the generel sentiment was that one needed to be on VS to be able to do that complex analysis.

    Anyone who has analyzed web sites for some time, knows that only a minority of visitors are truly engaged with the company’s proposition, and we are definitely in need of better metrics than duration, page views, returning visitors, etc., even segmented to the extreme.

    An interesting attempt is WebTrends new Engagement Index, which allows the marketer to give value points to certain types of action. I don’t know much about yet, though.

Add Your Comments


Print this Article

More articles from Robert Gorell

Marketing Optimization Blog
FREE Newsletter Sign-Up
send it once every: