First, spot the similarities. Over time, the strategies of incumbents tend to converge. A useful exercise is to overlay the business models of companies in the same industry and then look for areas of overlap. Wherever you see competitors doing the same thing, ask yourself, “What’s the shared assumption behind this policy or practice?” and then, “What would happen if we challenged that belief?” For centuries, innkeepers assumed you had to own rooms to offer guests a bed for the night. Airbnb inverted this belief and now has more than six million listings across the world.

Second, focus on what hasn’t changed. What aspects of your strategy have remained stagnant for years or decades? Over time, legacy practices, like wallpaper, become invisible. Your job is to question whether those 12 13 taken-for-granted practices still make sense. For example, though it endured a lot of pushback from traditional carmakers, Tesla challenged the long-held practice of selling cars through independent dealers. The company’s sleek stores, often located in luxury shopping venues, offer customers a hassle-free buying process. Tesla understands that the best orthodoxies to challenge are those that degrade the customer experience.

Third, go to extremes. Pick some parameter of performance—price, choice, availability, speed—and ask what would happen if we aimed for a 10X improvement? Fifty years ago, a retired physician, Dr. Govindappa Venkataswamy, launched an epic quest to eradicate unnecessary blindness in India. Millions of his compatriots had cataracts but couldn’t afford corrective surgery. How, Dr. V. wondered, could he reduce the cost of surgery by 90 percent or more? For inspiration, he looked at the fast-food industry. “If McDonald’s can sell millions of burgers,” he thought, “why can’t [we] sell millions of sight-restoring operations?” Today, Dr. V.’s network of specialty hospitals, the Aravind Eye Care System, performs half a million cataract surgeries annually.

I went to our board’s compensation committee and explained the dilemma. When you innovate, everything needs to change, not just the way you make or deliver a product. Many of the practices and structures within the company need to adapt, too, including, in this case, how the board rewards our executives. I proposed a radical idea— essentially, that I would determine compensation, based on how much they contributed to this new strategy, even though, without easily measured financial results, this was going to be far more subjective than our typical compensation practices.

I then proposed that, based on my reading, we had five ninety-day priorities:

• Stop hemorrhaging cash. We were precariously close to running out of money.

• Make sure we would be profitable in 1994 to send a message to the world—and to the IBM workforce—that we had stabilized the company.

• Develop and implement a key customer strategy for 1993 and 1994—one that would convince customers that we were back serving their interests, not just pushing “iron” (mainframes) down their throats to ease our short-term financial pressures.

• Finish right-sizing by the beginning of the third quarter.

• Develop an intermediate-term business strategy.

Finally, I laid out an assignment for the next thirty days. I asked for a ten-page report from each business unit leader covering customer needs, product line, competitive analysis, technical outlook, economics, both long- and short-term key issues, and the 1993-94 outlook.

I also asked all attendees to describe for me their view of IBM in total: What short-term steps could we take to get aggressive on customer relationships, sales, and competitive attacks?

What should we be thinking about in our long-term and short-term business strategies?

One of the most important resources a business can have is valuable privileged information—that is, knowing something that others do not. There is nothing arcane or illicit about such information—it is generated every day in every operating business. All alert businesspeople can know more about their own customers, their own products, and their own production technology than anyone else in the world. Thus, once Schultz initiated business operations, he began to accumulate privileged information.

Howard Schultz envisioned an Italian espresso bar in Seattle. He tested this hypothesis and found it wanting. But the test produced additional information, so he modified his hypothesis and retested. After hundreds of iterations, the original hypothesis has long since vanished, replaced by a myriad of new hypotheses, each covering some aspect of the growing, evolving business. This process of learning—hypothesis, data, anomaly, new hypothesis, data, and so on—is called scientific induction and is a critical element of every successful business.