Asking the Simple Questions
To people outside of data-related fields, the work that is performed by data professionals and the types of solutions that are built by them are often a bit of a mystery. I typically try to explain to anyone who expresses confusion or exasperation over not understanding what’s going on that data science - really any sort of data role - boils down to asking questions and trying to get answers.
And the type of questions that we’re asking, generally, aren’t a whole lot more complex than the types of questions that we learned to ask in kindergarten: Who? What? Where? When? Why? How?
One of my favorite sites to visit is Nate Silver’s fivethirtyeight.com, where they use a largely data-driven approach to follow the news of the day, from politics to sports. While they have some great reporting and some really novel data visualizations, in the end they’re just asking simple questions: Who is going to win the Democratic primary? Who is going to win the Super Bowl?
Their predictions, their answers to these questions, all boil down to working through a bunch of raw data, understanding the information it’s conveying, building knowledge about past events, and then using that knowledge to predict the future. Each step of that process is really just asking another simple question.
But one of the reasons that I believe that Data Fluency is an important skill for every member of any organization to have is that first step, of collecting raw data: because that’s really a process of asking and answering simple questions while going about the processes that make up a workday:
How many phone calls did I get today?
How long was each of them?
How many of the callers were upset?
What time of day did each call come in?
Was each caller a member of our loyalty club, or not?
One doesn’t need to have a Ph.D. in mathematics to be able to ask these questions or to get the answers; simply writing things down during the course of the day gets you that. But recording it - that’s collecting data - allows you to start asking different questions:
Has the number of calls I’ve gotten gone up or down lately?
Are the calls I have getting shorter, or longer?
Are more people taking advantage of our loyalty program?
While answering these questions takes just a bit more work to answer, it can’t be done without those team members who are working through your critical business processes, recording their observations. And they can’t know the types of things to observe and record if they’re not Data Fluent.