Today, Google pushed out into public beta, its Analytics API, which has been in private beta for awhile. This is excellent news on several levels. As with any API, access to data outside of its original, sometimes confining domain opens up several possibilities to the enterprising developer. End users can now develop reports and applications specific to their needs / requirements.
They can squeeze out more value from their analytics data by going beyond what Google already provides on their dashboards. They can also of course, mash up this data with other related datasets to gain a broader view of their operations.
I don’t need to enumerate the benefits of using an API. I am sure it is adequately and more insightfully covered elsewhere. However, I’d like discuss what this means to the analytics space as whole. Google changed the landscape by offering its analytics package for free. Granted, the pundits rushed to state the clear cut distinctions and limitations relative to its enterprise grade cousins, the value of Google Analytics, even in an enterprise environment, is now hard to ignore.
Yes, Google is used by a lot of small businesses who find the cost of enterprise grade analytics solutions prohibitive. In my humble opinion, the main advantages of enterprise grade analytic packages are;
- A higher degree of customization of tracking variables is possible
- Ability to define custom metrics
- Greater segmentation capabilities, and
- The ability to import and combine external datasets with clickstream data for custom analysis
With every upgrade, Google Analytics has inched closer in capabilities. What is cool about the API is that, all but the first item on the list above is now possible, I dare-say with even more flexibility than what is currently provided by others, simply because the sky is the limit to what you can do with your data. You can cut and segment it anyway you want. You can define new metrics without limitations. You can combine with other datasets for cross channel analysis. You can get into decision tree analysis with SAS to find that not so apparent insight. You can design a more holistic dashboard with a total view across all domains and a host of other things.
Note that point number 1 above is still a relevant and important difference. Imagine how Google Analytics compares to Omniture in the design of variables. Google’s main Achilles’ heel here is the provision of just one setVar variable for tracking any custom segment. The provision of multiple setVar variables will, in my opinion, completely negate the advantages enterprise grade vendors have. Other issues about ownership of data and privacy are less important at this point.
It is apparent that to fully leverage the API, you may have to use other tools to massage the data or write complex algorithms to create your segments and define custom metrics, all of which are already available in enterprise grade packages.
In the end, I suspect that many organizations will leverage in-house expertise to get more value out of this API. The ability to setup and use the API might also become a necessary skillset for budding Web Analysts. We are bombarded with massive amounts of data daily. The ability to parse and glean insights from this data, whether within an analytics package or outside of it, is ultimately what we seek. And in most cases, I suspect that working with the raw feed is the option that provides limitless possibilities. What say you?
Link: Google API News
Comments
No comments Posted. Start The Discussion!