Guest Post: Connecting Decisions to Data: A Case Study, Part 4: Decision Intelligence and Business Intelligence: How They Differ and Why We Need Both
In this post I discuss how Decision Intelligence differs from the current best practices in decision support epitomized by Business Intelligence solutions. My particular focus will be on why the typical “data-to-dashboard” Business Intelligence approach does not work for many if not most decision scenarios.
Please imagine that you’ve been given the decision model from the prior post from your analytics colleagues, who have done a good job performing the necessary research and number crunching to provide you with the information you requested. They might have used state-of-the-art data science (possibly including machine learning) to determine the demand curve and marketing spend to demand uplift relationships. They present the required information by creating a “Product Sales and Manufacturing Planning” dashboard on the company intranet, which looks similar to the example shown below:
With this data in hand, your task should now be simple enough: Can you tell me what are the right choices for levers 1, 2, and 3 below based on the information provided above?
- How much should I charge for each unit of the product?
- How many units should I order for my first run?
- How big an investment should I make in marketing the product?
(all of this should happen within the constraint of ensuring that the manufacturing run is profitable).
Look at the tables and charts for a while and try using them to answer the questions you will be asked by your executive review team. What is the combination of sales price, manufacturing run size, and marketing that will maximize the profit? How much profit will that result in? How would you use the above information to justify your decisions? How would you answer the question “How do you know there isn’t a different decision that delivers better outcomes?”
I’m guessing you might find it a little challenging to answer these questions, given this data. interested/qualified/committed
This scenario highlights why, in many situations, data science in conjunction with business intelligence alone doesn’t work for decision makers. The reason is a mismatch between what decision makers need and what data people provide. Data people often focus on providing, ‘insights’ from data’. The dashboard above undoubtedly offers many insights: those who are familiar with it will know more about the relationships that affect the manufacturing and sales of this product than those who are not. But these insights don’t help the decision maker very much. They do not shed any light on what outcomes will be achieved if certain actions are taken. That is, the insights provided by the data are not aligned with the actual decisions that need to be made. This is often not well understood:
In other words, even when data is available and contains sufficient information to solidly justify a decision, it may not be usable because it is not organized in a way that relates the different decision options to their outcomes.
So we are back to Problem 2 from my original post: How can we use these decision assets to show how decision outcomes depend on the actions taken by the decision maker?
To “re-organize” the data so it maps decision levers to decision outcomes, which is what is directly useful to decision makers—that is, to address Problem 2—we need to introduce the second tool in the Decision Intelligence kit bag: Simulation, the topic of my next post.
In addition to his duties as CEO, Zangari leads Quantellia LLC's Scalable Solutions division, where he is responsible for financial technology, telecom, and Covid-19 solutions. Zangari is also the architect of the company's World Modeler solution suite. Before joining Quantellia, Zangari spent 15 years as CTO for a spatial GIS company, where he specialized in providing solutions to utilities and telecoms.
A physicist by training, Zangari's papers on cosmology are still referenced today.