Few companies these days aren’t tracking measures and metrics for their Web sites. Some companies focus on raw measurements such as visits and page views; others report complex KPIs like conversion rates, bounce rates and visitor engagement. The technology to generate this data is pretty mature – Omniture, Visual Sciences, WebTrends and others are able to gather a rich set of data across the online channel. Visualization interfaces are increasingly easy to use, reporting options diverse, and customer support is, by and large, improving every year.
So why did 56 percent of respondents to a recent survey say they believed that Web analytics was “difficult?” And why did 69 percent of respondents in the same survey say they did not believe the majority of their co-workers understood Web analytics data?
Why indeed.
I believe the answer is simple: While there is a great deal of information about “why” you should do Web analytics and “what” you should measure, there is not enough information about “how” you make the data work for your online business. But the “how” describes the process of making good use of your current investment in Web analytics.
Too many companies have an unrealistic expectation about what is required to be successful with Web analytics. Here are three common mistakes most companies make with Web analytics and what they can do about it.
Don’t assume that everyone “gets” Web analytics data. Nobody studies Web analytics in college, and very few people have a strong understanding of the ugly nuances of Web data collection. Provide internal education, not just on the applications, but also on the data itself. Remember, knowing is half the battle.
Don’t confuse “reporting” with “analysis.” Web analytics is called Web “analytics” for a reason. Focus your efforts on producing data-based recommendations, not reports, and watch how people’s response changes.
Don’t spend all your time looking backward. Too many companies in my experience are looking for reports that describe “how they did.” Start looking forward by actively working to optimize your key performance indicators. And if you’re not running controlled experiments, start.
Educate, analyze and take action. Repeat as necessary.
