If you haven't been following the web analytics players these days, then you might not know that Microsoft if cooking up a free analytics tool to support AdCenter. Project Gatineau, as it's known, is currently in private beta. Microsoft's Ian Thomas took the eMetrics crowd through a fast demo and here's what I took away:
- Microsoft is pushing to key points of differentiation: attribution and targeting.
- Attribution is the question of which marketing tactic gets credit for a campaign. For example, if I see a banner today, search for chocolate tomorrow and buy on the keyword "hersheys" today, how do you credit the sale? From what I saw, Gatineau let's you input data about your campaigns and see all of the related metrics in graphs and try to visualize the interaction effect. This kind of visualization is great, but who gets the credit? First, last, equal distribution or some smart weighted average? If Gatineau intends to answer this question, I did not see it in the presentation.
- Targeting - Microsoft is gathering anonymous data about its users who sign into its Live service. Gatineau loops this data back into your analytics package so you can get some sense of the age, gender, and so on of your audience. But is this sample representative of the population? I wonder...
- Show Me The Data! - Data discoverability remains a key point of growth in the web analytics market. The more tools make it easy for you to find the data your looking for and, more to the point here, display data in a way that allows you to tease out insight more easily, the more likely you are to use and act on that information. To that end, Microsoft is offering 2 new features:
- The Timeshifter (my name) - you can easily look at data and quickly shift into a specific area that has data of interest. Thing of a big graph and a loop or magnifying glass that you can drag over it to get more detail
- Treemaps - Gatineau will permit the option of displaying content consumption through the use of treemaps (trust me, just look at an example over at Wikipedia and it will be better than me trying to write about it). Size represents popularity and color represents another metric of your choosing. This is not going to be particular surprising in aggregate (oh, people like my home page?), but if Gatineau lets you zero in on other details it will be WAY more powerful (oh, people love this product page, but never buy from it. Huh, maybe we should refresh it).
The 2nd version of the beta is getting ready to roll out. If you're interested in joining, sign up here.