These days, I’ve been spending more time around web metrics and analytics folks. These people apply technology to capture, measure, and analyze the data of site visitors on a website. The goal is to use all this data and analysis to optimize sites to get the best results.
At the most basic level, Google Analytics provides basic analytics on “bounce rates”, time on site, page views, click thru rates, and so on. More advanced products provide not only deep views into more detailed data on click data on sites, but also correlates the data to user segments and personas. Similarly, sophisticated testing products give the ability to perform A/B split tests, but also very complex multi-variant tests to determine the most effective combination of options to get the best results on a website.
There are other products that will analyze click streams to determine what web experience streams are most effective, and to determine the points where web experience streams break down. All this represents state-of-the-art of web analytics.
While I recognize the benefits of this “science” of the web, I’m also concerned about a potential bias or “blind-spots” that this focus on eMetrics creates. After all, there are humans on the other side of the mouse. And it is human behavior and motivation that makes them click. My concern is that the focus on analytics leads us to forget these human behavioral dynamics. At best, eMetrics and web analytics represent the symptoms of customer motivations. In other words, web analytics give you the effect, but they may tell you little about cause or human dynamics.
Often, the focus created by a preoccupation with web analytics leads to two approaches. One is to try to find places in a website that are broken. In other words, is there something confusing about a particular point of web navigation? Or is there a place that can be simplified to get better user outcomes? A second preoccupation is with trying various options to see what works best.
I must admit that many of these efforts of testing strike me as blind trial and error without much insight. It makes me think of the line, “Even a blind squirrel find a nut occasionally.” Even if this pattern of randomly testing option leads to better web results, I’m concerned that it doesn’t lead to fundamental insight. And if you don’t know why a particular configuration of web options leads to better website results, then it’s unlikely you’ll be able to gain insight and consistently engage the human motivations that lead to optimal results.
We must remember, it is humans that click, and human motivation and behavior is the ultimate cause.