It's useful to define paradox, as often used in management, in more detail. Leaving aside its literary uses, we define paradox, first, as a problem requiring leaders to recognize and fulfill two or more competing demands at the same time—for instance, winning at global and local levels simultaneously. You can single out what factors are under tension—and will always stay that way—almost like the strings on sash windows. In other words, paradoxes press leaders to serve one need while serving its opposite. That's hard enough on its own, but leaders are always facing multiple paradoxes at the same time and trying to find a way to proceed—when all the ways with a chance of improving current conditions are apt to seem illogical if not irrational when viewed from different perspectives.

Second, management paradoxes do not have a single right solution. They may have many solutions. Whereas in puzzles, two informed people will agree on a right and wrong answer, in paradoxes, they will rarely agree on either. The health care executive and her colleagues could agree on a strategy, because it posed a puzzle. But as it turned out, they could not easily agree on something else, because they faced a common paradox—how to create change in the organization while maintaining stability. More specifically, the challenge was changing people to fit the new business directory strategy. It turned out that she had to fire and hire to form a new winning team without disrupting the company's focus on performance.

Third, management paradoxes never get solved; they only get managed. Ron could choose to focus on just one competing demand—pressing ahead with a global marketing plan. But that solution would not—did not—last. In paradoxes, choosing to give priority to one opposing force does not eliminate the other. If you push for a global solution, local interests come around and hit you in the back of the head. If you push your solution down through the hierarchy by appealing to central authority, the decentralized authorities come from behind and trip you up if they don't agree on the solution. For example, if you get the CEO to champion your cause, the heads of functions, product units, and geographies who run the company can block your success. And so this year's solution, as the wheels of time turn, inevitably becomes next year's dud.

Solutions to puzzles are forever, rain or shine. Solutions to paradoxes must be monitored over time. They can change with the seasons. The difference between puzzle- and paradox-like problems—and the need to distinguish them—has gotten plenty of study both inside and outside business. As far back as 1973, Horst Rittel and Melvin Webber, professors at the University of California, Berkeley, described problems of social policy as either “tame” or “wicked.” Rittel and Webber referred to what we call paradoxes as wicked problems, in the sense of devilishly hard to resolve. They referred to puzzles as tame problems. And they noted that searching for rational, scientific bases for solving wicked problems was futile, owing to the difficulty of defining the problems, knowing correct from incorrect solutions, accommodating different perspectives of people involved, lack of definitive answers, and so on.4 “Proposed solutions” to paradoxes, they wrote, “are expressed as ‘good’ or ‘bad’ or, more likely, as ‘better or worse’ or ‘satisfying’ or ‘good enough.’ ” No true/false or right/wrong answer was possible, they argued. “The one best answer is possible with tame problems, but not with wicked ones.”

Even seasoned leaders get trapped into puzzle thinking, looking—as if in a multiple-choice quiz—for the correct answer. One CEO who had gone through many yellow legal pads filled with checklists told us in our research, “One day I looked at my list of problems to solve and realized I couldn't put a check next to any of the items on my list. Some of them had been on there for months, and I had one of those ‘aha!’ moments. I realized my list of problems were really paradoxes and they were never, ever going to be solved.”