Resources Articles of Interests
How much data do you really need?
Posted in Growth
Posted by Oren Harari in October of 2003
One of the biggest dilemmas for leaders is how much data they need to gather before making a decision. We know the dangers of "analysis paralysis" that mark risk-averse cultures (one client told me that his company was so risk-averse that "we study a good idea until it becomes a bad idea.") On the other hand, we also know the dangers of premature decisions that can lead to sloppy execution and catastrophic consequences. So what’s a leader to do?
While there is no equation or algorithm that provides a sure answer, I have found Colin Powell’s "P40/70" rule a very useful little formula to help leaders determine when evidence has hit critical mass and it’s time to take action. (Apparently, I’m not the only one who likes the formula. As I will describe in my next book, some pretty senior people in the U.S. Air Force and the Booz Allen Hamilton consultancy seem to concur).
So what is this mysterious "40/70" rule? In his 1995 autobiography, Powell proposed that if leaders believe that they have less than 40% probability of making the right decision with the amount of information they currently have---then they should seek more information. But how much more information should they seek? Here’s where it gets tricky. If military leaders decide to wait until they have sufficient information to hypothetically insure a greater-than-70% probability of making the right decision, they are likely to make the wrong decision because their decision will be too late. The battlefield conditions will have changed.
I would argue that the same applies to business decisions. When leaders insist on studying a problem until they have amassed sufficient information to be absolutely "sure" of making the "right" decision, they increase the probability of making the wrong decision—wrong not in terms of intellectual truth, but wrong in terms of competitive success. That’s because even while the organization is doggedly pursuing yet more data and information, the market battlefield often changes in a significant way. New competitors, new technologies, new value propositions, and new products emerge from elsewhere. Or, a fleeting, wonderful market opportunity is lost, and a nimbler competitor capitalizes on it. Either way, the organization’s ultimate "decision" is too little, too late, or too irrelevant. Moreover, I would argue that a corporate culture which insists on getting near-100% confidence before taking action is actually fomenting a hyper-cautious, cover-your-butt environment which will not only freeze entrepreneurial activity, but will also drive out the most talented employees. Ironically, procrastination in the name of reducing risk actually increases risk.
So what’s the solution? Powell argues "once the information is in the 40 to 70 range, go with your gut." Yes, you need hard data. You need solid due diligence. You need dispassionate analysis. Right up until the point you think you have a 40-70% probability of a good decision. But at the end of the day, you have to go with your instincts based on professional experiences and market knowledge.
Obviously, the 40/70 rule does not apply to the work of people whose activities inherently require as close to 100% accuracy as possible: people like auditors, scientists, engineers, software coders, quality inspectors, and the like. I apply the 40/70 rule to those who are involved with strategic, marketing, and organizational decisions—such as:
- Should we enter this new market—and if so, when?
- Should we outsource this function—and if so, to whom?
- Should we pour resources into this new product line (or this new technology)—and if so, how much?
- Should we seek a different kind of employee (or partner, or ad agency, or supplier)—if so, what kind?
- Should we overhaul our organization structure—and if so, how?
These are the sorts of decisions that beg for a 40/70 mindset.
On the surface, the 40/70 rule sounds irresponsible ("my goodness, how can anyone make a critical decision without having all the facts"??). Even worse, there’s no physics sort of precision as to when one attains 40% or 70% probabilities of success; it’s a judgment call. And yet, to successful corporate and military leaders, the 40/70 notion makes complete sense. Making an important decision with information that yields less than 40% probability of success might well be reckless. But the increments of information necessary to guarantee a greater-than- 70% success are increasingly costly to obtain, yield rapidly diminishing returns in terms of accuracy and usefulness, and are often counterproductive because they lead to slow decisions based on yesterday’s conditions.
The beauty of the 40/70 rule is that it constantly reminds the leader that the most optimal decision-making process is not an elusive quest for ultimate truth. Rather, it is a rolling process of intelligence and action: fast, conscientious intelligence-gathering followed by tries, experiments, pilots, project interventions, betas, and the like, all of which provide yet more data for the next round of action. You might call this the "Ready, Fire, Aim" approach to decision-making.
To summarize, the best decision for a leader is not necessarily the one that is the most accurate, but the one that is the timeliest and most likely to lead the organization to competitive success. Adhering to a super-ambitious P70-100% rule is quite appropriate for a structural engineer or a researcher in molecular biology, but it could be downright destructive for a leader of an organization. For leaders, who face challenges and priorities that are daunting, ultra-ambiguous, and filled with shades of gray, an ongoing cocktail of "40/70 data and gut" are the best predictors of success.