Earlier today, I read a good article entitled “The Art of Telling a Story In Analytics“. Not only did the author correctly identify communication as one of the most critical elements of analytics, but he also recognized the significant challenges that most face in that regard.
Consider the following simple construct of good story telling elements as it is applied to analytics:
1) Theme – The theme is what the store is trying to tell us. In analytics, this must be the actionable decision that will be made as a result of the analysis.
2) Plot – The plot represents the main conflict or struggle of the main character. In analytics, this should represent your hypothesis and supporting tests that you plan to execute.
3) Structure – The structure of a story defines the perspective (e.g. first person, third person) and tense (present, past). In analytics, structure should define the timeframe, duration, and context to be analyzed.
4) Characters – The characters are the heart of a good story, those around whom the events revolve. In analytics, the characters are the data elements. Like in a good story, they must be well defined and well understood. And usually many characters interact to produce a compelling narrative.
5) Setting – The setting represents the place and time in which the story takes place. In analytics, is is the same. But also consider elements of setting to include business performance, competitive positioning, or other elements of the business environment (e.g. merger/takeover, startup).
6) Style And Tone – These represent the language, words and actions that are appropriate for your story. In analytics, this represents the same, but with consideration to your brand positioning and desired customer experience.
By applying these elements to analytics storytelling, you’ll be able to gain the attention of your audience and explain the data and resulting recommended actions in an easy to understand manner.
I’ll leave you with a two important points below…
There is a very important difference between an analytics story and other stories. In a normal story, interesting is good. Interesting captures the reader’s attention, keeps them gripped to the pages and makes for an very enjoyable reading experience. In analytics, interesting does not help. As cited in the referenced article, knowing that a specific customer segment “tends to have many products, have not received many promotions , tend to buy financial products , are credit worthy and tend to live in Quebec” is merely interesting. It does not help drive a decision or business action. In analytics the Theme (decision) and Plot (hypothesis and tests) are critical to an effective outcome.
In a prior post, I wrote that “Data + Gut > Gut = Better Decision Making“, and this applies very well to the story telling construct presented above. Theme, Plot, Structure and Characters represent the Data; while Setting, Style and Tone represent the Gut. The data can only tell us so much about the story and point us in the right direction for the business decision at hand. But when coupled with the “Gut” components, representing the business context, the decision maker can be much better informed as to the optimal alternative. A story where we don’t know or understand the setting, style or tone can be confusing or misleading. But by adding these factors, the other elements of the story are grounded and make the story believable and hence helps to build trust in the process and decision…which of course makes for a compelling analytics story.