Does your organisation know what you mean when you talk about ‘bounce rate’? Do they know what a session is and what you really mean when you talk about a visit? I have just read an interesting article by Eric Peterson on web analytics standards in which he advocates using Google Analytics as a useful standard for web analytics definitions. I can understand his wish for standard definitions to work against, but is it really practical? Is it even desireable for most of us? I’m not so sure.
The problem with web analytics standards
The issue Eric is talking about is the fact that for every vendor, you have a different calculation of what constitutes a particular metric. For example: a ‘bounce’ in one tool may be a visit that lasts less than X seconds. In another tool it’s a visit to one page before leaving.
Multiply this problem by many vendors and many more metrics, and you can see it makes meaningful discussion across two different vendors very difficult. If you come to change vendors at any time, it’s very difficult to retain your benchmarks etc, as you’re not comparing apples with apples.
It’s unlikely the status quo is going to change any time soon, as vendors have a vested interest in their differences, or as they might call them USPs. If I have a one page site, heaving with content, does a definition of bounce that says ‘user only visited one page’ make sense? No! Every website is different, so it’s great to have a wide variety of choices out there that analysts can mix and match to create their own optimised solution. Why create web analytics standards that would make this more difficult?
Is it reasonable to assert that difference also drives improvement? Eric makes a good point that in the face of Google Analytics many vendors are going
the route of adding value to other digital marketing solutions. You may have read my post expressing disappointment at Google’s social analytics reporting, although it is improving. Omniture are now launching theirs. Whose will be better? I don’t know, but it will be interesting to find out, and the competition will drive improvement.
In short, I am doing my best to embrace the differences between different tools and get as much as I can out of them. As long as the analyst knows how his sessions are being captured, and analyses accordingly, there’s no need for a universal standard across all vendors. The specific problems this poses, like analytics books being wooly on detail, difficulty for professionals at conferences and workshops, don’t affect most of us. Analysts should make sure they know their own tool inside and out.
How important do you think web analytics standards are?