When presenting the outcome of analysis to stakeholders it is important to have a story you are trying to tell. This story has to make sense in a business sense, not just an analytics sense. Most of you hopefully know this instinctively, but for me it took a while before I fully realised both the importance of this and the implications. I was fortunate that I managed to have an innate curiosity that led me to mostly do this anyway, but now it is something I focus on when reviewing (or undertaking) any analysis. In this post I want to share two experiences from my career that cemented this view for me.
My first significant role in the private sector was for a mutual not-for-profit health insurance company. At the time I joined Australian health insurers had recently been allowed to be national organisations for the first time. The CEO tasked with bringing amalgamating the state offices of this insurer once said that when he started the only profit numbers he received were at a state level. My role was in finance and I was tasked with understanding the profitability of our products and customers. So I started by diving into the data and putting spreadsheets together to try to understand what was going on, but I soon realised I had a problem - although I could determine product profitability, I could not explain it. This led me to start spending time talking to various people with experience in the business and simply asking them what they thought were the drivers of profitability (there were numerous hypotheses), and on what basis they believed it (mostly company folklore). So then I set about proving these hypotheses as and when the data became available to test them. Over the time I spent in the organisation I was thus able to construct a story and proof around the drivers of profitability for the company. An aside - if you can help someone prove one of their pet theories or hypotheses it make it a lot easier to persuade them that other theories they have are false.
So that was my accidental story telling. Later in my career I was fortunate to work for someone who was very deliberate in this approach. She would collect all of the analysis done by the team and determine the story - then she would construct a deck slide by slide, only including something if it added something relevant and new to the story. This led to shorter decks than might otherwise been the case, and left me with one of the most important lessons an analyst needs to learn - just because you did all the analysis does not mean everyone else wants to or needs to see all the detail you went through to reach your conclusions.
So when you present you analysis, please keep it as short as is possible to demonstrate your point, and don't include things just because you did them or you think they are interesting - if they don't support your story or are repeating a point you already made, leave it out.