Telling the data story

Once upon a time there was a dataset…..got your attention, haven’t I? Well, the simple reason that the line is so timelessly catchy, is because it indicates the beginning of a story. And who doesn’t love a good story? It is said that hearing and telling stories is genetically programmed into our brain, as it helps our species survive and thrive by passing on customs, traditions etc.

Data however, is different. It involves a lot of math, statistics and some other fancy sounding words (algorithms, transforms, regressions etc) that turn off people. I have encountered this in my career- seeing people switch off when the analytics slides show up. And then it is up to me to make it worth their while to listen. A few tricks that have worked for me (and these are by no means original, as you will see) :

1. Make it real: For analytics to be interesting, it has to solve a business problem. I have written about this before, but this point bears repeating. More than just a business problem, it must solve a real, tangible problem that people can grasp. Think of (one of) the world’s favorite analytics specialist: Nate Silver. What gets people excited about the work he does is not the methodologies he uses, but his application. Predicting election results, winner of the super bowl etc-these are problems that people understand. And his work to solve them with data gets people’s attention. So to tell a good story, connect your story to a real problem that the client faces, and they’ll listen to you. (If you haven't heard of Nate, check him out here)

2. Respect the audience: There is an apocryphal story about a woman who got to go to dinner with two presidential candidates,X and Y. Upon completion of both the dinner engagements, she was asked who she would vote for-she said-Mr Y. When asked why, she said “When I spoke to Mr X, I felt that he was the smartest person in the world. However, when I spoke to Mr Y, I felt like I was the smartest person in the world”. Make your audience feel smart. Give them the feeling that they ‘get’ it. Without coming across as having ‘dumbed down’ what you are saying

3. Set the back-story: Give them the context. Why did you start this particular analysis? What sent you down this path? What was the ‘bad guy’ (Low sales, high acquisition costs) that you were trying to defeat through this analysis?

4. Emphasize the resolution, not the analysis: Speak about the findings and recommendations that will allow you to beat the bad guy. As I’ve said before-those findings are the ‘good guy’ of your story. Your analysis is not. At best the analysis is the tool the good guys use. And of course, make those findings actionable

5. Spare them the details unless…: You do not need to go into details of the analysis you performed. The audience does not care. They care about the implications and what it means for their work. However, fellow analysts do care about the details, so if they’re part of the audience, engage them afterwards to talk details.

Business Focused Analytics

I've been working in the field of digital analytics for the past few years. At my current organization, I have had some success in growing the analytics practice for three of our major clients. Starting off a couple of years ago, with a team of 1-2 people, we have a team of 12-15 working on those clients today. And I have learnt a few key lessons along the way. So here are a few ideas that have served me well:

1. Solve business problems, not analytical problems: As practitioners, we should remember that our end goal is to serve the business interest. We should ensure that our work adds value to the end goals of the business-whatever those may be. It is often tempting to lose sight of that fact, and wander down a deep analytical path that may be very interesting, but produces results that are not actionable in the real world. So before starting any analysis, ask the question: What do I intend to do with the results? And if the answers don't come through easily, reconsider your path.

2. The analysis is not the hero of your story, the business problem is : It is hard to imagine telling a story with data, but that's what we do, day in and day out. Effective storytelling is what turns great analysis into business transforming action. However, in doing so, keep the focus of your story on the business problem you're solving. That will make the analysis you did to solve it very interesting to the audience. As an example from Hollywood-would the action(the analysis) in the Terminator movies have been as interesting as it was, if the focus of the movie (the business problem) had not been preventing the ultimate annihilation of the human race by self aware robots? No one watches action for the action's sake.

3. To be effective, the analytics need not be complicated : Some of my most interesting insights have emerged from some of the simplest analyses that we have performed using Excel. In the age of big data, it is easy to imagine that terabytes of data will give us all the answers-and while that may be true some day-real insights lie not in the volume of data but in the art of effectively manipulating it. And with the right mindset, your tool set may often not matter.

4. Drop the jargon, tell them what it means for them : Data Analytics is complicated. There are a number of things that need to be considered, approaches to be used/discarded etc. But the end results should always be simple and easy to digest. I often think of our jobs being akin to those of physicians. They are the experts, knowing the intricacies of medicine, but when they talk to the patient-all the patient wants to know is "What's wrong with me, and how do you recommend I fix it". Leave the complicated jargon for the fellow analysts. Talk business with the business folks. And trust me, they will listen.