To my “Bias in BI” article a few weeks back, several commentors responded, questioning the relevance of the discussion for business intelligence. Though it was clear to me, I realized I hadn't done an adequate job prepping readers for my thoughts on the “science of business” that were behind the writing. I hope in the next two blogs to explain better just what that thinking is.
I wish I could take credit for introducing the science of business concept, but I cannot. However, I did steal the idea fair and square from MIT Sloan Management School researcher Andrew McAfee. Alas, my meager contribution amounts to little more than consolidating some of the thinking of business thought leaders such as McAfee and his MIT colleagues Michael Bikard and Charles Eesely with the evidence-based management doctrines of Stanford professors Jeff Pfeffer and Bob Sutton, the balanced scorecard strategic focus of Robert Kaplan and David Norton, and the analytics/experimentation directions of “Super Crunchers” author Ian Ayers along with “Analytics at Work” authors Tom Davenport and Jeanne Harris.
Pfeffer and Sutton provide plenty of motivation for the need of scientific or evidenced-based approaches to business in their award-winning book “Hard Facts, Dangerous Half-Truths & Total Nonsense – Profiting from Evidence-Based Management.” The authors catalog a number of poor management decision practices, the most notable of which is “casual benchmarking,” wherein companies cavalierly adopt what appear to be the best practices of industry leaders. The idea of benchmarking – using other companies experience and performance to set standards for you own company – is a good one. It's the casual part that leads to trouble. “The fundamental problem is that few companies in their urge to copy … ever ask the basic question of why something might enhance performance.”
Other heuristics that predictably cause problems are rotely doing what seemed to work in the past and following deeply-held yet unchallenged ideologies. The commonality of all three suspect rules of thumb: companies managing this way “show little interest in subjecting their business practice and decisions to the same scientific rigor they would use for technical or medical issues.”
McAfee outlines the scientific antidote to these dangerous heuristics of business strategy and decision-making: “… you have this unbelievable amount of horsepower and a mass of data to apply it to, you can be a lot more scientific about things. You can be a lot more rigorous in your analysis. You can generate and test hypotheses. You can run experiments. You can adopt a much more scientific mindset ... when you compare scientific to pre-scientific approaches, there’s one clear winner over and over.”
Bikard and Eesley expand on that thinking: “Viewing entrepreneurship through the lens of the scientific method is important for several reasons. First, it is important to understand exactly what hypothesis and assumptions form the core of the entrepreneurial strategy. This clarity allows the founding team to gain legitimacy more quickly since following this method illuminates the bet that is being made by investors and partners. Second, it enables the founding team to make sound predictions and judgments about ‘what is going on’ in the market and how to refute more efficiently or to validate the main hypothesis, thereby saving time and money ... in entrepreneurship, just like in science, explicit hypothesis generating, testing, and modifying is a paramount determinant of success. Aspiring entrepreneurs should keep that in mind, by fear of being selected out. In the 'Economic Experiment,' the failure rate is high.”
So the point of departure for the science of business is, not surprisingly, acknowledgment that perhaps the best approach for developing, testing and evolving business strategy is the same one that's been driving scientific discovery for centuries:
- Observe a phenomenon or group of phenomena
- Formulate an hypothesis to explain the phenomena. The hypothesis takes the for of a causal relationship or a mathematical function (the more of X, the more of Y).
- Use the hpothesis to predict the results of new observations.
- Perform tests (experiments) with the predictions.
Part 2, to follow next week, will put more detail around the science of business and draw the linkage to BI and analytics.