Web analytics, the impressionist portrait of your user experience.
Are web stats picture-perfect? Turns out they're more like a Monet than Minolta.
An attendee from the NetSquared Twin Cities presentation asked me if it’s true that Google Analytics (GA) doesn’t do a perfect job of tracking new users on your site?
A coworker had explained to her that because GA uses browser cookies, which are occasionally reset, users who have previously visited the site, but have recently deleted their cookies, will show up as “new.”
It’s true, and the question highlights an important misconception about web analytics. As it turns out, there are any number of reasons why the statistics in Google Analytics are not exact, ranging from problems related to tracking cookies, to multiple users using the same computer, to javascript blocking and more.
As Brian Clifton points out in his book on Advanced Web Metrics, “Too many businesses take web analytics reports at face value…The harsh truth is that web analytics data can never be 100 percent accurate, and even measuring the error bars can be difficult.”
Ouch.
So what then? Delete your tracking code and call Lady Cleo to reveal the secrets of your user activity? Well, no. As Clifton explains, although errors are there, the errors also tend to be constant. In other words, the number of people who are tracked inaccurately as new users, as in our example above, tends to be roughly the same month to month. Although your numbers are not precisely accurate, the trends themselves are still relevant.
That’s why when I talk to people who bring up this issue to me, I explain it this way. Although GA is not perfectly accurate, it still has the power to give you important insights about how effectively you’re communicating to your audiences about your cause.
It’s very much like viewing your user behavior as if it were an impressionist painting. The alternative being not seeing anything at all.
Be aware of the inaccuracies absolutely, communicate to your ED and board the difference between trends and exact numbers yes, but stop using analytics? No. The cost to your ability to make better decisions for your donors and constituents is simply too great.
[...] Cary Walski gives her impression of Google Analytics’ accuracy [...]
[...] Cary Walski gives her impression of Google Analytics’ accuracy [...]