Can you tell a good story? Three secrets to telling a good analytical story
At the end of the day, we use business analytics to tell a story. A story to our customers, a story to our peers, a story to our managers. A research story of a business phenomenon we’ve discovered, hypotheses we’ve evaluated, interesting findings, and explicit conclusions. Even the simplest of BI solutions, helps us dig for some facts and surface a handful of findings, which we can stitch together into a story. The better solutions will draw charts and diagrams for us, helping us create colorful, convincing presentations. The handful of real high end business analytics solutions will lead us to a valuable, useful story – finding correlations, causation, insights. This starts to be quite interesting. There’s tremendous business value in a story like that.
Of course, at any level of analytical solution, the story is only as good as the data that goes into the system. If the data isn’t up to date, the story won’t be very relevant, will it? Only a few years ago we were still willing to work with BI systems that worked in batch mode and came up with findings that were long history by the time they were “discovered”. Who is willing to compromise like that anymore with today’s advanced, near real time systems? For a story to be relevant in needs to be fresh!
Another problem with some solutions lies in the sources of data they use. If the data that comes into the system is limited, and doesn’t integrate different sources, our story will be quite limited as well, and more often than not, it will be skewed. Again, only a few years ago, memory, bandwidth, and CPU power were still costly and many solutions were designed to look only at a slice of the organization. This created a silo effect – important correlations ignored, assuming wrong causations and leading to misleading insights. The business value of such partial solutions is very limited. Unfortunately, in many organizations, analysis is still done in data silos, not leveraging granular enough data.
Traditional analytical platforms were not designed for a collaborative exploration of data. Rather, they assume an individual contributor conducting an analytical research and later sharing her findings. However, with today’s business complexity a multifaceted analytical research is so much more valuable. A solution that fosters a dynamic dialogue between various constituents in the organization will bring in a variety of insightful views and gain immediate buy in from all contributors.
Today’s challenges are different than those we faced at the early days of business analytics. We have oftentimes more data than we know what to do with and the challenge is finding a solution that can pinpoint the real insights in this “big data” and help us tell a relevant story that drives business value. We can no more afford to rely on dated, limited, or one-sided information. Making decisions based on non-relevant information is worse than using no information at all. As Stephen Hawking said, “the greatest enemy of knowledge is not ignorance, it is the illusion of knowledge”.