Path Analysis: A Paper and Some Thoughts.

I just wanted to quickly touch base on a topic from an article I recently reviewed for the Web Analytics Association Peer Reviewed Journals project: “Path Data in Marketing: An Integrative Framework”.

This paper was appealing because it made me think on all the data and information that is being gathered by organizations and how much (or how little) of this information is actually being used. I agree, may be a lot of it is noise and can even be useless. There are even pieces of information that should not be collected at all. For sure we are missing to track many golden nuggets.

However, so many valuable insights can be hidden on those data stores. In that context, how valuable would it be to dig more into path related data?

Here is a quick summary of the article:

“Executive Summary

By analyzing different marketing scenarios or “domains of data collection” (grocery shopping, eye tracking, web browsing, information acceleration), the authors attempt to formally define a path in the context of marketing. Their main objective is to propose a “unifying framework” that will facilitate further research on the topic. This is accomplished while showcasing the importance of path data in the future of marketing research.

The authors define paths as “records of consumer movements in a spatial configuration” or as “a conscious agent’s movement in a physical or simulated environment that is observable”.

As part of the suggested framework, two primary dimensions are identified: “characteristics of spatial configuration” and “the agent”. Specific characteristics of the actual spatial configuration, either if it is physical or nonphysical, continuous or discrete, as well as the existence of constrains, will influence the model and the approach of the analysis. Defining agent factors such as the level of (1) “social interaction”, (2) ”goal-directedness” and (3) “forward-looking-behavior” should also be considered.

Based on literature review, references to relevant related research sources are introduced. Several examples in the context of “retail/service environments”, “advertising studies”, “e-commerce”, “experimental research” and areas outside marketing are also considered. In the next section important operational issues, considerations and restrictions for researchers working with any form of marketing path are identified.

Finally, the authors conclude on potential directions of future research in the field.”

You can find the complete review at the Web Analytics Association Website.

If you are a member of the Web Analytics Association you can even get a single copy of the full journal. To request a copy, email Lindsay De Santis.

[...]

Why will this become important? Well, may be path analysis can answer some critical business questions that we are not able to solve now and we are just guessing at the moment.

Look for example at what Next Stage Analytics is doing. They claim they are able to predict demographics and detailed visitor information, among other projects, by only looking at their behavior. I wonder, how much of their approach relies on some sort of complex path analysis?

I feel that a deeper analysis of this data, and more specifically of path can become more and more important in the months and years to come. Who knows, I may be wrong. Am I?

[Added - August 27th 2009 ]

If you found this enticing you can also check some interesting thoughts and comments on the topic at Christoper Eyes on Web Analytics Blog.