A Data Science Central Community
We all get used to looking at Business Intelligence as a tool for strategic decisions mostly for top management. In looking at today’s business intelligence software market you will find a wide array of various types of dashboards, reporting tools, and KPI monitoring tools to support decisioning at the top levels. We have also seen a few good BI applications for decisioning on the front lines, supporting such vital areas as Sales, Risk Management, Budgeting, and other functions. We can definitely say that a Top-to-Bottom approach is actively developing over the time.
Well, that is natural for many reasons. Top managers make high-value decisions. At the same time, they are both budget holders and budget users which enable them to quickly grasp all possible benefits from business intelligence software usage, and then promote the approach at the departmental level after the success is proven at the middle and upper management levels.
But what’s on the other hand?
Looking closer at the BI industry as a whole we can see that the very nature of such an approach prevents further business intelligence solutions development and potential. Following this “Top-Only” approach companies limit the possible effects from the BI usage. In the majority of cases, it is unreasonable to build data transformation and forecasting models, as well as identify KPIs and support correct data collection, and ultimately supply reporting under constantly changing market conditions for use in the trenches.
In practice, it is much wiser to develop tools which support the organization from the bottom up. Having a number of high-value but seldom supported BI generated strategic decisions, you would still take into account a large number of day-to-day operational decisions that are necessary during the course of the organization’s regular activity. This would include aspects such as customer acquisition or cross-sale decisions, to name a few.
For these kinds of decisions an absolutely different approach should be applied. Instead of crafting a dashboard it is better to give a clerk a single figure – an easily understandable form linked to an action. For example: potential profit for a new customer over the next 12 months divided into simple categories: “negative”, “low”, “normal”, “awesome”, with clear actions linked to every category: “reject”, “accept”, “assign to junior account manager” (or automated processing), “accept and offer better terms” (to avoid loosing an awesome client to another vendor). Such an easily understandable system can save your organization the same millions as a strategic decision about entering a new market or changing pricing policy.
The Bottom-Up approach suggests identification of all operational decisions that can be made faster or more accurately using data we have about our customers and results of our operations.
Like in our personal life, when we tend to think that our decision about the next car defines our quality of life and happiness; underestimating every-day decisions, such as what time to wake up or what to choice to make for lunch; we shouldn’t forget that these everyday decisions can make us much more successful and fulfilled than any new Jaguar.