It’s an age-old question. Why do successful technology companies stumble once they reach the top? Legendary businesses such as Digital Equipment Corporation, Polaroid, and Atari all led their industries yet eventually succumbed to the kinds of technological innovation they had pioneered. The pace of such change is quickening. In 2008, RIM’s BlackBerry conquered the mobile-phone industry, and the company’s value hit $80 billion. Less than five years later its devices have become passé, and its stock has tumbled 90 percent. Today Microsoft, Nokia, and even Apple must be wary of similar fates.

Developing breakthroughs involves creating something new and valuable (see “50 Disruptive Companies 2013”). It is inherently uncertain. And therein lies the challenge. As organizations become successful and grow, uncertainty is the enemy. They seek to eliminate variation and increase efficiency. They identify best practices and design standard operating procedures. This can make a business wildly efficient at what it does today. But it has a serious downside: an avoidance of novelty that can eat at the very soul of a company.

To overcome this challenge, companies must relearn skills and capabilities that brought them industry leadership— adopting processes that create variation, not eliminate it, and valuing flexibility over the relentless pursuit of efficiency.

First, they must design diversity into their DNA, to combat the natural forces that act to reduce it. For example, Google allows some engineers to spend 20 percent of their time on side projects and gives added resources to those ideas with the most merit. Intel funds hundreds of university grants to learn about new technologies outside its “silicon road map.” Netflix tapped the creative juices of hundreds of external developers, offering a $1 million prize for the algorithm that could best predict consumers’ movie preferences.

Second, organizations must embrace change, not resist it. They must explore, without expecting the initial specification to represent more than a starting point. Their processes must reflect the fact that your first plan is always wrong—meaning the challenge is to quickly and cheaply work out how it can be improved. They must understand that trial and error allows rapid discovery of what cannot be uncovered through even the most brilliant analysis and planning.

Ultimately, staying ahead in a digital age requires companies to welcome the uncertainty that endangers their current business. Only by doing this can they discover the new possibilities that lie ahead.