Right-mixing Science and Art for better Decision Making
Decision making today is more oriented towards data, metrics, predictability models and controls. Risk department forecasts their numbers using heavy duty econometric models. Many of these models are based on absolute precision. Recall the phrase ‘show me the data’ before making any argument. The point is organizations are heavily depending on science today to run the business. We are not challenging science here; it is the way it is being used to make decisions. Why are we consistently getting our numbers wrong? Now there is news about China’s real estate bubble and rise in US foreclosures. Why we make wrong decisions based on certain system and do not learn from them? Why we go back to the same system after failing? Using sophisticated financial models, CEO’s make next quarter guidance only to get them wrong again and later get criticized for not delivering on their promise.
Even the biggest conflicts of the world cannot be solved with high precision weaponry. It is the art to use the force and the table to resolve issues. It is time that organizations realize the need to mix science with art to make better decisions. Models alone cannot take the blame for failing the system. Many predictive models are based on the assumption that future behaviour is based on the past. There is a need to qualify this assumption. Art is the ability to apply science for making better decisions. The ability to test the models taking cognizant of the assumption and limitations can lead to fewer errors in decision making.
Just by doing trend analysis, seasonality tests, or scenario analysis cannot make a model robust. Justifying the upper and lower limits, understanding the quality of data used and the time dimension would help to converge and understand the scenarios better. Measuring quantity is easier compared to related quality. Quantity provides a status of a metric but doesn’t tell us anything about customer experience, sales quality, competitive landscape – these are constructs which needs a different scale to understand, measure and act.
To understand the scenario better, we need to supplement the quantitative models with an artistic mind. We need a larger canvas for the two to engage, verify and conclude. Yes, if it cannot be measured, it may not be important but it could also be that it is the most important problem we are facing and it is just that we have not figured out how to measure it.
