I recently presented a keynote presentation at the annual conference of the joint Ohio chapters of the Institute of Management Accountants. I shared the stage with Mike Willis, Partner, PricewaterhouseCoopers and the founder of and Chairman Emeritus of XBRL International. Mike posed this question to make an important point:
Two trains are on the same track 600 miles apart traveling towards each other. One train is traveling west at 50 mph and the other train is traveling east at 20 mph. How long will it take for the trains to impact each other?
The narrow correct answer is B. But an experienced analyst thinking out-of-the-box would ask (1) Who would ride on a train traveling 20 mph? and (2) Who would put two trains on a collision course?
Framing a problem
In my mind there are two pre-requisites to problem solving: (1) first frame the problem or opportunity, and (2) then perform the analysis.
Framing a problem is usually not an easy task. Yes for simple plans, it is. For example, one decides to take an umbrella if the sky has dark clouds but not if it is sunny. Is one 100% sure? It is probably good enough for the umbrella decision. But do you know or just think you know?
This example gives a glimpse of the limits of planning. Mental shortcuts, gut feel, intuition and so on typically work except when problems get complex. When problems or opportunities get complex, then a new set of issues arise. Systematic thinking is required. What often trips people is they do not start by framing a problem before they begin collecting information that will lead to their conclusions. There is often a bias or preconception. One seeks data that will validate one’s bias. The adverse effect is we prepare ourselves for X and Y happens. By framing a problem, one widens the options to formulate hypothesis.
Analytics begin with a hypothesis
Ah, the term hypothesis. Posing a hypothesis critical and requires analytics, the second pre-requisite, to prove or disprove if the hypothesis is valid or not. Much is now being written about analytics. There is a reason. The margin for error keeps getting slimmer. Also, once accepted types of strategies (e.g., low-cost producer) are vulnerable to competitor actions. The only truly sustainable strategy is to have organizational competency with analytics.
Experienced analysts frame, analyze, and then plan for actions. But plan to re-plan – numerous times. And today with high performance computing (HPC) there are ways to do this. Reliable forecasting and probabilistic scenario planning can be added to the portfolio of analytics that progressive companies use.
I love this example. There’s so much to talk about, I’m not quite sure where to start!
So I’ll just babble for a bit.
Who would get on a train going 20 miles per hour? In the real business world, there are LOTS of companies going 20 on various ‘trains’, when they need to be going 120.
Who would put two trains on a collision course? Lots of companies measure (and compensate) one department on a KPI that is on a direct collision course with the performance goals of another department. So the ‘success’ of one department directly creates big challenges for another department, all within the same big team.
In this latter example, the measurements are ‘framed’ in a way that doesn’t take the big picture into account. This organization is analytically incompetent!
As for planning for action, can anyone do anything to keep the trains from colliding in 8 hours and 34 minutes? To whom will the analysts give this information in order to save lives?