Great Escapes
Nine decision-making pitfalls--and nine simple devices to beat them.
By MICHAEL USEEM

(FORTUNE Magazine) – All in all, our decision-making equipment is pretty sound. We don't follow the lead lemming over a cliff. We can't be fooled into thinking that a 99-cent lure is a meal. We don't try to catch car fenders with our teeth. Then again, it wasn't a dog who launched New Coke. So there are a few bugs--little design flaws of the mind--that can have big consequences. People are clinically overoptimistic, for instance, assigning zero probability to events that are merely unlikely (such as a massive iceberg in the path of a really big ship). We see "patterns" in the random movements of stocks the way our ancestors saw bears and hunters in the scatterplot of the night sky. We make choices that justify our past choices and then look for data to support them. Not only do we make these errors; we make them reliably.

That's the good news. Predictable errors are preventable errors. And a few simple techniques, like the nine below, can help you steer clear of the most common wrong turns.

âñ PROBLEM

Analysis paralysis

âñ DEVICE

The 70% solution

Good analysis is a necessary precursor to every decision. But only professors and journalists get paid to say, "On the one hand.... " Managers are paid to decide. So when they continue to mine and massage data in the pursuit of perfect knowledge--and thus perfect certainty--they're succumbing to what psychologists have dubbed "decidophobia."

The Marine Corps battles this syndrome with the "70% solution." If you have 70% of the information, have done 70% of the analysis, and feel 70% confident, then move. The logic is simple: A less than ideal action, swiftly executed, stands a chance of success, whereas no action stands no chance. The worst decision is to make no decision at all.

Maintaining decision "tempo," as the Marines put it, can require deadlines of one's own making. During General Electric's "workout" retreats--where managers convene to pinpoint what is pointless--the ranking manager is required to say which of the many proposed changes he or she would accept or reject. Forcing a thumbs-up or thumbs-down on a decision ensures it won't suffer endless analysis.

âñ PROBLEM

Sunk-cost syndrome

âñ DEVICE

Burn the boat

Seymour Cray built two things: sailboats and supercomputers. Each new Cray supercomputer--he produced his first in 1963--was its own masterpiece, crafted with all the care and brilliance of a Stradivarius. One even had a decorative fountain in its coolant system. But in computing, Cray knew, there is no such thing as timeless perfection, only obsolescence. To drive the point home, legend goes, he'd build a beautiful sailboat each spring, then burn it in the fall.

It's always painful to destroy something we've built, whether it's a machine, an organization, an idea, or even a paragraph. Our investment is both psychic and economic. But the cliché holds: no pain, no gain. "My advice to young men," Henry Ford wrote in the 1920s, "is to be ready to revise any system, scrap any methods, abandon any theory if the success of the job demands it." Ford did just that--until the systems and theories in question were his own. Presented with a prototype to succeed his Model T, one story goes, Ford literally took an ax to this effrontery.

He was axing his company's future. By the time Ford introduced the Model A in 1928, General Motors had stolen his lead. Cray, on the other hand, was sinking his sunk costs.

âñ PROBLEM

Yes-man echoes

âñ DEVICE

Voice questions, not opinions

"I don't want any yes-men around me," movie mogul Samuel Goldwyn once said. "I want them to tell me the truth, even if it costs them their jobs." Ha, ha, ha. We laugh because it's true. Organizations aren't in the habit of rewarding people who speak uncomfortable truths. And yet they need to hear them.

"If you walk into a room as a senior person and innocently say, 'Here's what I'm thinking about this,' you've already skewed people's thinking," says Marine Gen. Peter Pace, nominated to become the next chairman of the Joint Chiefs of Staff (see his interview in "How I Make Decisions"). His approach: "Start out with a question and don't voice an opinion."

Why? Because people can't line up behind you if they don't know where you stand. And if you present subordinates with an intellectual challenge, they feel freer to offer their opinions without fear of giving offense. "If you are looking for answers, ask the question," advises Pace, and "if you are looking for an honest critique, you ought to be the first person to self-critique." It's different from generals' metaphorically taking off their stars to ask for honest feedback. People know the stars will be back. And besides, says Pace, "I like my stars."

âñ PROBLEM

Anxiety overload

âñ DEVICE

Look at the clock

Low levels of anxiety are productive: They concentrate the mind. High levels of anxiety are counter-productive: They panic the mind. A panicked mind stops processing new information, reverts to tried-and-true responses, and is prone to snap decisions that make things worse.

What you need is a circuit breaker to snap the anxiety loop. Steering a faltering F/A-18 toward a carrier deck, some Navy pilots steady themselves by studying the clock. Various gauges may be spinning alarmingly, but the clock is not. Emergency-room physicians check their own pulse. Firefighters lay a hand on the shoulder of an anxious colleague. Calm, like anxiety, is contagious. So devise something that works for you. Maybe it's walking to a quiet corner. Maybe it's listening to the sound of your own breath. Maybe it's picking up the Bible. These things are in a steady state and can help put you in one.

âñ PROBLEM

Warring camps

âñ DEVICE

Let the battle rage

In the early 1980s, Gillette had an internal war on its hands. It was losing market share to Bic's plastic throwaway razors, and the question ricocheting through the company was "Are you a steel man or a plastics man?" One camp wanted to meet the price competition head-on by pushing Gillette's own plastic disposables. The other wanted to cede the low ground to Bic and invest millions in creating better metal razors. Many CEOs would have stepped in to quell the strife. But this was a decision Colman Mockler couldn't afford to get wrong. So he let the divisions slug it out.

Political infighting can be destructive, but battles over substance, managed well, can be constructive. A furious volley of fact is met with a fierce counterattack of analysis--and the battlefield is littered with useful information. The self-interest on both sides tends to cancel out as long as the boss stays neutral.

For nearly two years Mockler played a perfect Switzerland. Then one day he walked in and promoted the champion of the steel camp. Its argument--which would lead to "shaving systems" like the spring-mounted Sensor, the Mach 3, and razors marketed like jet fighters--had carried the day.

âñ PROBLEM

A wily adversary

âñ DEVICE

Clone your opponent

In January 2004 football's New England Patriots were facing a uniquely dangerous opponent. Peyton Manning, quarterback of the Indianapolis Colts, couldn't be contained by conventional means. His arm was too accurate, his feet too quick, his style too different. So a week before their big game, Patriots coach Bill Belichick gave his backup quarterback a challenge: Become Peyton Manning.

The backup, Damon Huard, did more than study a few game films. He threw himself into the role. "I read about him," he told the Tacoma News Tribune. "I tried to duplicate his mannerisms as he got to the line, how he called his audibles, how he handles the ball, all the little things I could come up with." Huard's performance, teammates said, was Oscar-worthy--and the real Peyton Manning suffered for it. Given a week to scrimmage against Manning's doppelgänger and to adjust their tactics accordingly, the Patriots intercepted Manning four times to advance to the Super Bowl. After the game Belichick singled out Huard, who didn't set foot on the field that day, as a key contributor to the win.

Cloning translates easily to other competitive situations. Since nobody's better at spotting holes in a plan than an opponent, assigning a person (or a group) to think like your competitor can expose flaws that, identified early, are less likely to be fatal. By making it their job to stress-test a plan--much as new products are dropped from heights and set afire--in-house clones can guard against overoptimism without dampening the energy needed to carry out a plan. Nobody wants to be tagged a naysayer. But imitation can be a high kind of heroism. Just ask Huard.

âñ PROBLEM

To be? Or not to be?

âñ DEVICE

Go with the omen

Yes, a prophetic sign. When our rational brains tie us in knots, our natural preferences sometimes express themselves in nonrational ways.

In 1993, Bill Clinton's Trade Representative, Mickey Kantor, asked Charlene Barshevsky to serve as his deputy. Here was a chance to help shape U.S. trade policy at a time of rapid globalization. And here was an agonizing choice. Barshevsky was a partner in a prominent Washington law firm, co-chaired its international practice, and earned several times what she would receive in government. Moreover, she was married and the mother of two small daughters. Would joining the Clinton administration be the right decision for her and her family--or a dreadful mistake? She became "paralyzed," she said.

A full week and a half had gone by when Barshevsky resolved to make up her mind that day--though she was still irresolute. Driving into Washington with her husband, she told him that she was out of time, that she wished she had a "sign." Several minutes later she saw a car with a vanity plate: "Go4It." "I'm going to say yes!" she exclaimed.

Barshevsky was unconsciously ready for a trigger to get to yes--and when she saw it, she went with it. Her mind was made up. She just hadn't realized it.

âñ PROBLEM

Inexperience

âñ DEVICE

Educate your instincts

"Go with your gut." "Follow your intuition." "Trust your feelings, Luke." We've all heard the sayings. But do your instincts make good decisions?

It depends on their education level. The instincts of veteran city firefighters have been shaped by years of experience, so when something tells them to stay off that staircase, there's usually a good reason. Less experienced firefighters perform worse at fire scenes, research suggests, because they lack such unconscious triggers. Untempered by a sense of how bad things happen, they are propelled by native optimism to blunder ahead.

Blind instinct cannot be trusted--but it can be educated. The main purpose of flight simulators, for instance, is to make pilots experience unlikely surprises so many times that, should one actually occur, their response will be reflexive. An after-action review accelerates the process by squeezing learning out of each experience. "If you get educated about something and then you live that, the line blurs between what your instincts used to be and what they are now," says Gen. Pace, citing this as one of the reasons he decided to earn his MBA. "Your mind touches on resources it's not even conscious of touching on."

âñ PROBLEM

Self-interested thinking

âñ DEVICE

What would Sara Lee do?

Even when we want to, it's hard to separate a company's interests from parochial ones. Wall Street has its own agenda. Employees have theirs. Lawyers and auditors aren't always disinterested parties. And neither are we, even when our decisions are supposed to transcend private objectives.

There's no simple way to navigate this thicket of competing, even conflicting, interests. But here's a useful compass: Imagine, for a moment, that the company is a person. If Sara Lee were an actual woman, that is, or General Electric a retired Army officer, what decisions would they make for themselves? Like most people, companies want to live long and prosper. They want security, growth, good relationships, respect, and a sense of purpose. Since companies can't speak or decide for themselves, their people--from the CEO on down--have to do it for them.