The Golden Hat

With the release of free web analytics tools like Google Analytics, and the impending release of similar analytics tools from likely suspects such as Microsoft, the everyday website owner is bound to get less sophisticated. And so will most companies.

Why?

Because “free” typically translates to “easy” for most people and “easy” means…… easy. The user enters their analytics dashboard to encounter hundreds of different reporting metrics and functions and despite the daunting stockpile of web analytics, we somehow believe that our job is done as a website producer.

How many visitors are checking out my homepage and my product landing pages? How many visitors are bouncing off of my new product page and which links are they navigating to instead?

Let’s assume for example that 60% of visitors are exiting a particular product description page without making a purchase whereas the average exit rate of other product description pages is about 40%. Should you safely assume that people aren’t interested in that product and remove it from the top level product pages? What actionable decision should you make? Perhaps the site visitor is VERY interested in the product yet they found the same product on your competitors site for 10% less. Which actionable decision should you then make?

My point is that installing web analytics and looking at the raw numbers is not enough. It only tells you a snippet of the story. If analyzing site traffic were a novel, installing analytics and looking at the raw numbers is chapter 1 of a 15 chapter book. The number 15 is an arbitrary number but I think you get my point.

Most companies fall into the same trap of assuming their job is complete post installation of analytics, falling into the same lull of counting daily, weekly and monthly unique visitors, sourcing out a few referring domains for link building purposes and calling it a day. Yet web analytics is much more than simple domain and landing page traffic counting. Many of the reporting functions in even the most popular web analytics packages are pointless, such as hits (which has long been retired as a useful measurement). As Avinash Kaushik also points out screen resolution is another pointless and distracting metric reported by most analytics tools. Over a several month period you’ll notice this metric rarely ever changes. Why not look at industry reports instead, such as through comScore, and identify before site design or re-design what resolution your industry visitors would prefer?

As a company, or even as an independent site owner hoping to build their online persona, I recommend thorough research and training on web analytics. Omniture is one of the prominent analytics providers that also offers an analytics training and certification course for their analytics tool. Typically if you become an expert with one of the major analytics tools then you’ll be an expert with the others since they tend to provide the same reporting capabilities. The major difference between them is their dashboard interface. Yet one should not only know how to navigate the dashboard but how to make actionable inferences from the data they see.

I’ll have more analytics post to come. Server log data, web analytics, PPC data….. fun stuff

Like this article? Share, save, and spread it!