When Aaron Baker arrived at Harvard last year, to begin work on web analytics, he had…Google Analytics, and its data. But that wasn’t really enough to do the kind of reporting that their office needed. While they had data, Aaron really needed to be able to tell stories.
Before you can turn data into stories, you’ll need to get a lay of the data landscape. Where are you? Where do you want to go next? What’s your “situation”? You’ll need context to be able to interpret your data in a way that will let your stories emerge. “Data is what, but it’s not always why.”
As an example, Aaron shared the trend of mobile hits to Harvard’s news site growing closer and closer to desktop views…until recently, one viral news story actually pushed mobile’s numbers higher than desktop. What story did that tell? That it’s really past time to take a mobile-first approach to website development.
Typical data analyst questions include:
What are business objectives? (Context)
How do we support them? (Relevance)
Where is the data stored? (Access)
How was it collected? (Procedure)
Any relevant acronyms from the above are left as an exercise for the reader.
Defining your situation and telling your stories requires a few steps. First, gather all the data you can — not just website analytics, but data from email campaigns, social media accounts, and multimedia repositories (such as YouTube). Plot them on a Cartesian graph, with near/far on the X axis (how easy is it for you to get access to the data?), and light/dark on the Y axis (how well do you understand the tool and its data?)
After you’ve drawn up this data landscape, it’s easier to set goals for your data collection and your stories. Do you have a lot of Google Analytics data, but not a great understanding of how to use it? Are you a whiz with your email campaign system, but are your open rates and clickthroughs hard to get from your provider?
Once you’ve gotten a better handle on your data and your stories, it’s important to share them. If you’re publishing 75 news stories daily (as Harvard is), there’s room for a daily analytics roundup. But you may want to keep the basic analytics about page views and demographics to a quarterly report, which could be used to tell a story with other data.
Share your stories, backed up by data, with the people creating your content, because that feedback can help them do their work better. And of course, you never know when the boss is going to have that ad hoc request.
Finally, if you’re at Harvard, you might even put up a large display in the lunchroom that shows a constant current stream of social media activity and data. Because nothing says lunch like data.
Digital Analytics Lead, Harvard University