As the instructor for the Spring 2014 Introduction to Web Analytics class at the Community College of Vermont, I look for examples of “websites (or analytics) in the news” to combine learning the basics of web analytics, which frankly can be dull at times, with something more interesting and relevant to real life situations. The current semester began this past January, and the launch of the Vermont Health Connect website was very much “front page” news; the idea for this assignment began to emerge.
Here’s a quick example of how a snippet of java script can take you from boring code to important insight in just a few clicks (this is not unique to the VHC site, all websites can be viewed in this way):
Go to the Vermont Health Connect home page:
View the source code (How do I do that?) to see:
Focus on the Google Analytics tracking code, boxed in red, which is this bit:
Simply recognizing that this tracking code is in place helps determine that from the outset, Google Analytics was built into the website, which is great. That means that even if one only uses the standard built-in reports, one can analyze all pages on the site to determine many things, such as how many visits the page has had, the flow of the page views (which exposes bottlenecks or problem pages), how searchers are finding the site (shows keywords used), and what other websites may be referring people to this one. With some configuration you could track, how many people completed signing up, and if they did not, where they exited the site. If both the website and the analytics were configured correctly, those analytics reports could provide data on how many currently uninsured people are completing the signup process, and it can show data in real time.
This is where I think things can get interesting. According to a Vermont Digger article:
“The Shumlin administration is not tracking how many uninsured Vermonters are signing up for Vermont Health Connect. Statistics on the uninsured won’t be available until the next household insurance survey. There is some disagreement about when that survey will be conducted.”
This is an excellent example of a point on which class discussion could begin. Why not track that data? It is possible, through some good design and analytics tracking. The tracking would be anonymous, so it would not show who is uninsured. Instead, a well-designed signup process could include a box one must check that might say “no current health insurance,” or “currently insured,” etc. and those check boxes would be counted for completed signup processes. Privacy, therefore would not be violated if the site tracked those numbers. That same tracking code in the image above tells us that whoever has access to the Google Analytics account UA-33924458-3 for the vermont.gov site has data now, and can tell what is being tracked, and use that to inform the web designers (and the governor) to make changes that would then enable better tracking of the data being collected. I would also be sure to point out in class that web analytics data cannot be the one source of numbers for the analysis of any actual website’s performance or traffic. Other data collection methods used in concert with web analytics data are required for actionable data-based decision making for the business or service behind the website. Vermont still needs the survey data.
The assignment to the class is to look at the Vermont Health Connect site and apply the analyses that we are learning about in the course to that site. The intent is to demonstrate how understanding web analytics tracking enables you to look at a site and offer insights on a website’s performance in relationship to the website’s purpose. We started with the keyword reports on Google Trends, and we continue to look at the site as we get further into our exploration of analytics reporting. I hope you will take the time to read some student posts and join the conversation by leaving a comment. You can contact me by email at email@example.com.
Thanks for visiting.