Changing the Odds
Why 90% Right Often = 0: How to Improve the Odds on Your Most Important Project Growing your business depends on new projects—creating new products, entering new markets, finding new customers, and acquiring new operations. Likewise, improving your business depends on major new projects, such as upgrading your IT system, simplifying your manufacturing processes, and streamlining your supply chain. But at the moment of launching any project, there’s a problem that most of us don’t come to grips with: the inherent, all-too-human tendency to be overly optimistic about the odds of success.
What are the odds that the new venture your company is about to launch will produce a marketable product within the next eighteen months? What are the chances the merger your company is about to conclude will create rather than destroy shareholder value? Out of twenty new products now in development in your lab, how many will be on the market two years from now? And of those, how many will be profitable?
Even approaching such questions in a spirit of objectivity is very tough for most people. As researchers Daniel Kahneman and Dan Lovallo explain in their article “Delusions of Success,” “We chronically overestimate, to a dramatic degree, the odds of project success.” We look at a project with a 5 percent chance of success and we think the odds are 30 percent; we look at a 10 percent chance and we think that it’s 50 percent.
It’s natural for businesspeople to look on the bright side when estimating their chances for success. Optimism generates energy. Those who are chronically pessimistic attract few followers and accomplish very little. But optimism has an Achilles’ heel: It causes you to overestimate the true odds at the outset. Take a minute and think about the true odds in most common business scenarios. A close look at the data suggests that the failure rate for many types of projects is in the range of 60–80 percent. (See Figure 1-1 for typical failure rates for specific project types.)
Typical Failure Rates for Specific Project Types
Company merger or acquisition60%
Information technology project70%
New food product78%
Venture capital investment80%
New pharmaceutical productOver 90%
In truth, every project you undertake is a kind of suspense story. How and when and where will it go wrong? What unexpected obstacles will arise to derail it? There are many, many ways a project can fail. They range from undercommunication among team members to conducting too few experiments and considering too few options when designing the new product or the new business; from relying on a flawed technology to using a technology that works but costs too much or takes too long to develop; from failing to anticipate a competitor’s preemptive move to inaccurately forecasting consumer demand; and from overlooking the need to retool your marketing infrastructure to support a new product to ignoring the time bombs planted by internal politics that will blow up any chance of successful implementation.
This list of traps and pitfalls is far from complete—you can probably extend it dramatically based on your own experience. Consider all the ways that a project can fail and you may find yourself wondering how it is that any projects succeed.
Of course, many projects do succeed. Some IT programs work phenomenally well; some new products become enormous hits. But too often, excessive optimism impedes project success by underestimating the investment and the sense of urgency needed to drive the crucial de-risking moves. The people in marketing are consulted only twice rather than ten times about the details of the CRM project; the new product reaches the market six months too late, with a vital feature or distribution channel overlooked and omitted. Time, energy, and treasure are wasted, and the expected growth breakthrough never happens.
The first step in changing this picture is carefully estimating the true odds—overcoming the natural human tendency to be overly optimistic with a bracing dose of realism and data. The purpose is not to demoralize the effort. It’s to give you a clear, accurate sense of what’s really needed to make your project succeed—how much investment is needed (usually twice as much as you were planning), how many smart moves you have to make, how many contingency plans you need to prepare, and how smart a business model you need to design.
Projects live in a tough, probabilistic world. Even if you do everything right, the odds can turn against you. And if you misread the odds and underinvest, you can guarantee that the project will fail. There are so many ways to underinvest—not just financially but also in terms of time, energy, emotion, resilience, options considered, conversations held, experiments conducted, simulations run, market trials conducted, tires kicked, and doors slammed. Making matters worse is the fact that for many types of business initiative, success is an all-or-nothing matter. Unless all the elements required for success are in place (the right technology, the right value proposition, the right customer set, and so on), the project can fail—completely. This little-recognized reality of project risk can be described by the simple formula “90% right often = 0.”
Thankfully, there are proven techniques for changing these odds for the better. The fastest way to learn about them is by studying the companies that pioneered them. For example, the Japanese automaker Toyota created several of the most powerful of these techniques while developing the Prius—a blockbuster product whose odds of success were less than 5 percent when Toyota embarked on its development in the fall of 1993.
Toyota Changes the Odds on the Prius
Step inside the minds of the leaders of Toyota in the early 1990s. The company was riding high, enjoying growing market share and unrivaled profitability. Any manager in this situation might have been tempted to be complacent or even arrogant.
But Toyota’s executives responded to success differently: They were extremely worried. They’d seen how complacency had weakened many other successful companies, including the great American automakers. They feared that, despite their current success, a feeling of product maturity was setting in. Hungry, cost-conscious, hyperefficient competitors such as Korea’s Hyundai were emerging, eager to do to Toyota what Toyota had done to Detroit’s Big Three.
The interior monologue of Toyota’s leaders at this inflection point might have sounded like this:
We’ve created great breakthroughs in quality, pricing, and fuel efficiency. These have made us the world’s fastest-growing and most profitable car company. But what about tomorrow? What is the next great breakthrough in our industry? How can we ensure that we will create the breakthrough, rather than be victimized by it? What must we do to anticipate the risks that success will bring and transform them into opportunities before they blindside us?
You might call this “competing in advance,” a strategy for preempting risk by outthinking it, much as chess master Garry Kasparov plans his fifth move partly to fend off the forking attack his opponent might launch with his seventh move. Competing in advance may not be a formula for personal serenity, but it’s a powerful tool for success if you are trying to protect your company from its biggest risks.
In an effort to make sure that Toyota would remain in the forefront of the auto industry, the company decided to mount an all-out effort to create the first great car of the twenty-first century, almost a decade before that century would arrive. In 1993, a team of ten Toyota board members met to imagine the qualities that a breakthrough car (which they code-named the G21) for the coming century should have. They envisioned a car that was comfortable, safe, pleasant to drive, appealing to female motorists, low-polluting and environmentally friendly, and highly fuel-efficient—sound ideas, but amorphous. An engineer named Takeshi Uchiyamada was assigned to convert them into a concrete proposal.
Uchiyamada, a specialist in techniques for eliminating noise and vibration, wasn’t an obvious choice for the role. He had never headed a new-vehicle development team. But from 1991 to 1993, he’d led a task force that reviewed Toyota’s R&D process from head to toe. This turned out to be crucial preparation for the G21 challenge. It exposed him to many parts of the company and deepened his understanding of the disparate technologies that Toyota had developed. He had become an expert in the inner workings of the company, its many strengths and its less-obvious weaknesses.
He had also learned where the company’s most talented engineers were. Now he recruited ten outstanding engineers representing all of the key technologies that would go into the G21: the body, the chassis, the engine, the drive system, production technology, and so on. All were in their early thirties, old enough to have experience but young enough to be flexible.
They worked for months to give a concrete form to the G21 concept. They then brought their proposed car to the company’s executive vice president of R&D, Akihiro Wada. Uchiyamada explained that his team hoped to create a small sedan that delivered 47.5 miles per gallon, about 50 percent better fuel efficiency than the Corolla, a comparable current car. At the time, it must have seemed to Uchiyamada an impressive proposal. He probably presented it with no little satisfaction, even a bit of quiet pride—which would have made the reception he received doubly shocking.
Wada listened politely, then replied, “Fifty percent better is not enough. Our competitors would quickly overtake us. Please double it.”
Uchiyamada was taken aback by Wada’s challenge. “At that moment I felt he demanded too much,” is how he later put it in his understated manner. This was the first in a series of blows that would have shattered the psyches of some less-resilient managers.
Shaken but determined, Uchiyamada returned to the drawing board. He realized that the higher performance bar meant that he would have to rethink his assumptions about the G21. Tinkering with existing technology wouldn’t suffice. It would require a major leap to an unproven system that existed only as a blue-sky concept Toyota had been studying—the hybrid engine. Even today, Uchiyamada shakes his head when he thinks about the nature of the challenge: “There were no models, no comparisons, no benchmarks. Everything had to be created from scratch, with no guarantee that any of it would work.”
What’s more, Toyota wasn’t the only company considering the hybrid engine. Everyone in the car business had heard rumors about hybrid experiments at Honda and Ford. How far had they advanced? When would they announce a breakthrough? No one knew, but the existence of competing projects must have sharpened the anxiety Uchiyamada felt when he realized he’d have to commit to a hybrid model. It wouldn’t be enough to create a successful new design; he also had to do it faster than the competition.
Given the realities, the G21 project was a long shot. Takehisa Yaegashi, who joined the Prius effort as project leader, put it this way: “We didn’t think it was impossible, but the odds were very low. Perhaps 5 percent; perhaps even less.”
One in twenty. Would you bet a billion dollars on those odds? That would be a very tough call. But the people at Toyota didn’t merely put their chips down and spin the roulette wheel. They made a series of moves to change the odds.
Recognizing the complexity of the technical problems the company faced, Uchiyamada needed to find new ways to surface problems and solve them quickly. He started by creating a new system designed to facilitate communication and joint problem solving among his team members.
The system started with a dedicated physical space. They called it obeya, the big room. It was equipped with personal computers and two computer-aided design workstations. Team members were asked to assemble in this room daily to work together on the G21 project—the first time this had been done at Toyota.
Uchiyamada also created a virtual space, an electronic mailing list compiled to encourage team members to quickly and broadly disseminate key issues and problems as they arose. Over time, this list grew to include three hundred people.
Of course, e-mail wasn’t a new technology at Toyota, but the way it was used on the G21 project was new to this relatively hierarchical, formal company. Here’s how Ikujiro Nonaka, a professor of strategy at Hitotsubashi University, described the process in his article on knowledge creation at Toyota:
“Equal access to information” was one of Uchiyamada’s action guidelines for the G21 project. . . . In an ordinary product development project at Toyota, when an employee found a problem he reported it to his boss. And if the problem couldn’t be solved, it would be reported to the chief engineer, who would notify other engineers who could be affected by the problem. This was a time-consuming process. In the Prius project, by contrast, the engineers could post an e-mail to the mailing list immediately after discovering a problem. Anyone who read the e-mail and had a solution could immediately post the necessary information.
The visual metaphor for this kind of communication is not a pyramid or even a network but rather a sphere, as in the old definition of God attributed to the fifteenth-century mystic Nicholas of Cusa, a sphere “whose circumference is nowhere and whose center is everywhere.” Everyone on the mailing list stands equally close to the center of the action, and everyone is capable of being the center at a particular moment in time—able to draw energy from everyone else in the group to solve today’s most pressing problem.
Excerpted from The Upside by Adrian J. Slywotzky with Karl Weber. Copyright © 2007 by Adrian J. Slywotzky. Excerpted by permission of Crown Business, a division of Random House, Inc. All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.