The Google Analytics Decision
Most of you know that my favorite price for things is "free." (I’ve even put together the Skinflint Internet Marketing Guides for those that want to spend nothing for their campaigns.)
I’ve long used Google Analytics on my site as a free way to track how many people are coming, and have recommended it as the cheap way to do the same on your site. In December, Google announced a new version of Google Analytics, prompting the question of whether you should switch.
Whether you should switch to the new Google or not depends on what you have now.
If you already have Google Analytics, you will eventually need to switch to the new version. Google has announced that the old "Urchin" script is being replaced by the new "GA" script and the old script might stop working by the end of this year.
My site was using that old script, so I spent about 20 minutes today changing to the new script. I had used a technique called a server side include to place the script on every page of my site, so I just needed to change that one file to make the change. I can’t tell how well it’s working yet, but I’ll look at my numbers over the next few days to see.
The tougher question is what you should do if you have a different metrics service or software package—one that you currently pay for. In the past, a few features have driven most adoption of high-end metrics:
- Better reports. Google Analytics has reports that handle most situations, but some companies depend on customized reports for their business that go beyond what Google can do.
- High volume traffic. Google Analytics has been alleged not to be able to keep up with very high volume Web sites, relying on sampling to randomly collect data. Dropping data wouldn’t be everyone’s favorite way to solve this problem, however.
- Tracking more than page views. Until now, Google Analytics could not track events other than new Web pages. Now it can track Flash, AJAX, and other events.
- Data sharing. Some suspicious types worry about Google having access to their metrics, so they buy someone else’s solution instead. No feature added to Google Analytics will quell this crowd.
My take: You won’t do better if your budget is zero. If you already use another metrics solution, you must evaluate not just the loss of function, but the pain of migration and the disruption in your data. No two metrics systems count things the same and they can’t ingest the old system’s data for you to see your history in one system.
I’d be interested in hearing from others who have reasons to use Google or one of its competitors for metrics.