In customer experience management, we talk a lot about silos: data silos, information silos, organizational silos, and so on. We often silo ourselves into one category or the other: I'm a numbers person or I'm a words person. I am on the data science team or I am on the marketing team. I'm a salesperson or I'm an engineer. However, the idea that data analytics is just for numbers people is not only wrong, but also incredibly limiting.
Think about it this way: If you're in the business of data analytics, then you're in the business of storytelling. Storytelling, though, isn't just about moving narration or crafting fictional pieces. It's about using data to take the audience on a journey that compels them to act or think differently. Today, some of the best storytellers we have are in the field of investigative journalism; we can borrow some of their techniques to learn how to use data more effectively.
Investigative journalism fundamentally differs from more conventional journalistic styles. The latter works to deliver the facts by answering the key questions around who, what, where, when, why, and how and delivering an objective, truthful vision of the world. The former, on the other hand, uses objectively true material toward the subjective goal of reforming the world. This approach requires questions that dig a lot deeper than just the who, what, when, where, why, and how. Investigative journalists do intensive research to gather clues and evidence; this data is then used to construct compelling stories that persuade the reader to action.
If you're working in CEM, this paradigm should sound familiar to you! Your data likely points to the things that need to change, but you're often met with reasons why it can't be done. No budget, shifting priorities, no available resources; the list goes on. If this saga sounds familiar, it's likely your story is too conventional and needs to get a lot more investigative. The solution lies in what I like to call "investigative analytics."
Like all good mysteries, investigative analytics begins with gathering initial facts and clues. Rarely, though, are these clues enough to tell the whole story. By organizing and thinking critically about the clues, you might find gaps in logic that could erode a water-tight narrative. To close these gaps, we must ask thoughtful questions that interrogate the data further until we're confident we can refute all doubts about the story. However, it is important to note that for that story to be powerful and persuasive (and thus, get the investment you need to make change), you need to be asking questions that dig deeper than just the facts.
Our brains are not wired to remember facts for long periods of time, but they are moved by skillful storytelling, so we need to focus on asking questions that will help build an irresistible story. To do this, start with the basic questions, but then refine them so that their answers can lead you to a deductive conclusion about your story. For example, rather than asking "When does our satisfaction score drop?" you can ask "Which day of the week do our customers typically struggle the most?"; or "Are our satisfaction scores unequally affected by certain holidays?" Rather than asking "What topics are discussed most?" or "what is the average KPI?" you could be asking "What combinations of topics are customers talking about?" or "What is the polarity of comments within a topic?" or better yet, "What is the relationship between customer effort and Net Promoter Score, and are there any outliers?"
Asking questions in this way leads us to far more interesting insights than asking questions for which we typically already have answers. One of the biggest issues with asking basic questions is that the information they point us to is not actionable and is, therefore, largely ineffective. When you look at the analytics maturity curve, the first ledge is descriptive analytics, which is as far as we can get with those basic questions. Further along the curve, however, we find predictive analytics. Many of us have struggled to make the leap from descriptive to predictive due in part to stale analysis techniques. To cross this chasm, you must start by asking better questions of your data.
The goal of this exercise is to make your data tell a story that matters. In the last year alone, we've seen very real examples of the way investigative journalism has sparked entire social movements and led to important changes (think #timesup and its relationship to Ronan Farrow's writing). While your data might not offer the kinds of insights that change societal structures, it is very likely holding the key to stories that reshape your company, its products, or the way your customers experience your brand name. You just need to ask the right questions so that you can tell a prize-winning story.
Ellen Loeshelle is a principal product manager at Clarabridge.