Multiple Measures Means Better Web Analytics

An interesting post on the Metrics Insider by Josh Chasin of ComScore describes the potential conflict of two forms of online measurement tools, panel based and site based. I agree with Chasin that the question should not be which data source is “right”, since both have their strengths and limitations and neither method can lay claim to absolute accuracy. The question should be to look at both types of data and ask why they may bring one to a similar conclusion or to a divergent one on a particular analytic issue.

For those who have worked in the CPG industry with both traditional retailer store scanning data (IRI, Nielsen) and their panel data services, you know that you can have two data sources looking at similar events (product purchasing) and see differences and still get value out of both. You have to know their roles and limitations, and you need to triangulate them with other sources to get the best answers.

I realize from an advertiser or portal standpoint, they’d like to focus on just one form of measurement and preferably one they have some influence over, but I agree with Chasin that in general we should all really “take advantage of this embarrassment of riches”.