INFORMATION TECHNOLOGY

Home > IT > How bad decisions get made

---

How bad decisions get made

October 24, 2007

What was the worst decision your organization ever made? Launching a product too early, or too late? Getting too big too soon? Pushing through an unpopular and ultimately failed program?

Drug giant Merck would no doubt cite its decision to keep the painkiller Vioxx on the market. The company stands to leak profit for years if even half of the current lawsuits related to coronary risks are successful.

So what led to this disastrous decision? And how can you avoid such fateful decisions in your own organization?

The canon on this theme runs deep, but for our brief analysis let's start with some decision theory.

Three stages of decision-making

Decision-making involves three stages: identifying possible strategies to achieve a goal; weighing the pros, cons, risks, and constraints of each strategy; and then choosing what seems the best.

Individuals can accomplish this in a split second, while groups may spend months in deliberation, with equal amounts of success.

Of course in reality it is we complex and conflicted humans who are making the decisions. In the murky pre-decision air, countless aspects of our psychology come into play and interfere with the process.

We make decisions based on tradition, without data, or despite data. We throw good money after bad. We don't see the forest for the trees. We want to get ahead. We seek to avoid public disgrace. We want to please the boss.

And on and on.

When the smoke clears, the bad decision always seems so obviously flawed that we cannot see why that option rose above all others. Why, for example, did invading Russia, vast and cold, seem like a good idea to Napoleon?

Decision-making is a process, not an event: Harvard Business School

Harvard Business School professor David Garvin urges us to view decision-making not as an event, but as a process evolving over time and involving many people.¹ This was certainly the case at Merck, as the court documents reveal in painful detail.

"Sometimes decision-making processes get their own momentum," says Garvin.

"So you can't always tell who made the decision or when. Instead, the weight of evidence accumulates, people start to snuff out their doubts, and the process moves to a seemingly pre-ordained conclusion."

As Garvin describes it, any organization's decision-making steamship could hit ice in due course. Here we offer three suggestions for clearing the fog and avoiding that iceberg decision.

#1: Know your culture

Organizations have their own tacit decision-making culture.

While your organization may have many healthy decision-making practices, most groups exhibit at least one if not several symptoms of a dysfunctional decision culture.

Such unsound tendencies might include squelching dissent, delegating decisions upward, neglecting to decide until it's too late, obsessively gathering data, spreading decision-making power too widely, or being unwilling to speak out against the majority.

As in psychology, where recognizing your illness is step one toward a cure, so it is with recognizing the defects in your decision-making culture.

Look carefully at your flaws, socialize the analysis to achieve general agreement, and ensure people are trained in recognizing your decision-making pathology.

With general awareness of and better transparency in the process, more people are better positioned (and more willing) to intervene when the process jumps the rails.

#2: Imagine the worst

One of the strongest inhibitors to rational decision-making is "mob rule."

Most organizations – consciously or unconsciously – reward team players and discourage naysayers. This pressure may come from above and/or from peers, and can manifest itself in the extreme as intimidation or threats.

Merck's own Vioxx critics experienced both, says Prof. Garvin.

A powerful ally of mob rule is self-censorship: people imagine themselves alone in their doubts or questions and keep quiet.

You can work against this natural phenomenon using what professor Garvin terms "constructive dissent."

To achieve constructive dissent, formally assign a team (it's too hot a seat for an individual) to find flaws in potential strategies, different views on an issue or product, or new options altogether.

Giving voice to dissenting views, even if it's part of an exercise, lets the decision-making process surface multiple viewpoints, alternatives, and unthought-of outcomes.

Your possible solutions will broaden from "yes" or "no" to "what about?"

Less formal constructs can include getting leaders to remain mum on a topic until everyone has spoken, fomenting dissent not just among the ranks but also in the executive boardroom, and rewarding candor.²

Could all of this negative feedback instill the fear of failure and paralyze decision-making? As an antidote, follow imagining the worst with imagining the best.

#3: Let the data guide you

The data present in your organization can help you learn from your past to take the right step forward.

As long as you can trust the information, you should give it a fair audience, though it may be counterindicating your instincts.

While it is primary to trust your information, it is also critical to know which data is essential to your decision. Be warned that it's not always the most obvious choice.

In the Vioxx case, information may have failed to flow to proper levels of organization. For example: the commercial organization may not have been clearly connected to the clinical and regulatory teams. In addition, there may not have been a clear benefit to being the person who identified the risks and brought them to light.

The lessons of Moneyball

In his book Moneyball, Michael Lewis provides a great example of how to reach a goal using the appropriate, though non-traditional, information.³

Web Demo

Performance Management

In this case, a low-budget major-league baseball team, the Oakland Athletics, won games against vastly more expensive teams because of the way they picked their team.

The management phased out old-style scouting concerned with highly visible talents such as "hose" (throwing arm), left-handedness, and an athletic body. (High points in these realms spelled an expensive draft.)

The new approach looked at less immediately appealing traits such as fewer walks and more strikeouts for a pitcher, more walks and on-bases for a batter. Looking for these unsexy stats helped them find cheaper players who would consistently win games for the club.

The A's set their strategy and used the data to carry it out.

Data: too much of a good thing?

It's true that we sometimes look at the wrong data. We also look at too much data.

One of Malcolm Gladwell's conclusions in his book Blink is that people make the best decisions after considering only a subset of relevant data, which is not to say all of the available data.⁴

To wit: medical researchers looked at thousands of pre-heart attack exams and found that the presence of only four symptoms clearly indicated that the patient was having a heart attack.

Other symptoms such as age, stress, and smoking seemed relevant to the point of being critical, but were in fact not good indicators.

It is important to mention that researchers came to this conclusion only after studying heaps of evidence. If this is not possible, an alternate strategy is to leave certain data out of a less important decision and analyze the effect.

If the data seems to have no bearing on the outcome, perhaps it is not a key factor.

Summary

Organizations cultivating a performance management culture that encourages fact-based decision-making stand to learn from the mistakes of others.

Even if corporate decision-making is made simple by the overarching guide of increasing shareholder value, we have seen how a bad decision can unwittingly erode this goal.

In your case, a bad decision may not lead to the decimation of the French army, but you'll still feel the pain of the mistake.

In a process riddled with human failings, it helps to cultivate a fact-based decision-making culture starting with complete and reliable information.

On that foundation you can build practices for the appropriate use of data and the treatment of unhealthy decision-making tendencies.

And remember our three keys: know thyself, appoint contrarians, and listen to the data.

Find Out More Like this article? Check out similar ones on decision-making Discover the information sweeet spots that lead to better decisions Order your copy of The Performance Manager

---

Sources

¹ Garvin, David. "How You Company Can Make Better Decisions." HBR Podcast, July 5, 2006

² Emmons, Garry. "Encouraging Dissent in Decision-Making." Harvard Business School Press, Oct. 1, 2007

³ Lewis, Michael. Moneyball. W.W. Norton & Co: New York, 2004.

⁴ Gladwell, Malcolm. Blink. Little, Brown & Co: New York, 2005.