Feb 2nd 11 Posted by CJ Cunniff in Analytics, Digital, Featured Articles, Marketing, Research

Remove the Data That Doesn’t Matter

As promised from my last post I’m going to talk about something that sounds very “advanced” but is actually one of the easiest things to understand in the world of analytics.  Advanced segments are a tool that Google has created to allow us to view a subset of all of our data.  This is a very important topic for us to understand because data in aggregate tends to be one of the root causes of something called paralysis by analysis.  So let me lay out a scenario to help us understand the problem and then we can look at the solution and, more importantly, how this can help the higher education industry.

Let’s say that we notice that a particular page, which was designed to generate leads, has a long average “page view” time, but it isn’t generating the leads we need.  A natural question would be, “What is causing this phenomenon?”  To answer that question, we will need a few more pieces of information, such as:

1. Where is the traffic coming from?
2. How many pages have these people viewed on our site and what is their average visit time (in other words, are these engaged audience members)?
3. Who is the target audience for the page, and which audience is actually viewing the page?

There are plenty of other questions we might need to ask, but you get the point.  Now, to get to the crux of the issue, you may be wondering how in the world you are going to get some of this information.  Here are a few simple suggestions:

• If you know that you have specific marketing pieces being sent out to certain areas (cities, states, etc), you could create a custom segment for visiting traffic by geographic region where those cities or states are included.  The result will be that you can see if the same pattern exists among those key users.

• You can create a segment where the users have visited the site more than once, and have a visit time of more than a minute.  Now you will see if there is a pattern among the more engaged audience members.
• One additional idea is to look at search terms used to get to your site (I would recommend a segment for both your internal site search and outside searches from services like Google or Yahoo!).  This will help you understand more about what the users are likely looking for and, with certain search terms (working adult programs) for example, you can even get an idea of your audience makeup on the page.

Some quick tips for executives:

1. Remember that Analytics will virtually never give you an exact answer or perfect information.  However, the goal is to give you a reasonable amount of certainty in making decisions that would otherwise just be gut decisions.
2. You need to be viewing at least an overview of your website data regularly (once a month at an absolute minimum).
3. When you are viewing the nifty charts and graphs that your web or marketing department is sending you, ask yourself a few basic questions, like:
   • Are there any peaks or valleys in the data that stand out?  If so what could be the cause of those?
   • Has this information been segmented properly to give me a realistic representation of the audience segment that the site is intended for?
   • If the information has been segmented properly, then how can we further optimize our content to meet both our users’ needs and our institutional/site goals?

Now it’s your turn:

What types of segments have you found to be the most effective and why?

What’s next?

We can certainly glean a lot of valuable information from our clickstream data (Google Analytics Data), but for those topics or reports that can have a significant effect on marketing decisions, we need to make sure we aren’t misinterpreting the data.  So, in my next post I’ll talk about tools and methods you can use to combine research with your analytics data for a solid understanding of your audience(s), their experience on your site, and what you can do to improve it.