Everything from historic compensation structures to relationships with component and manufacturing partners to sales disruption channels gets in the way of the mistakes required to inspire disruptive innovation. Why else would so few Fortune 500 companies—just 5% by one estimate—be acting decisively to embrace digital technologies that are disrupting the business models of entire industries?

It’s a phenomenon that Box (BOX) CEO Aaron Levie has dubbed the “Industrialist’s Dilemma,” the subject of a graduate course he taught last winter at Stanford University’s business school.

“The next decade or so will be defining for which industrials will survive and which won’t,” Levie predicted.

One recent example that illustrates his thesis: Tesla’s (TSLA) relatively muted response to controversy over a driver fatality believed to be linked to its autopilot technology. “If you have a Detroit company had that sort of failure, that would halt all innovation” until the matter was resolved, Levie said during a main stage session Tuesday at Fortune Brainstorm Tech. “The cost of failure is more public.”

Yet, a willingness to make mistakes is exactly what big companies must embrace to avoid being disrupted, especially by software specialists that can speak directly to their traditional customers. “Failure can set a positive example,” Levie posited. “You will invest a lot of money in things that don’t work.”

That’s a philosophy that General Electric (GE) is trying to instill throughout its culture from managers to engineers to sales representatives, said Beth Comstock, vice president of industrial giant GE. “We’re trying to be both digital and industrial,” she explained. “To be big and yet act small.”

For the past five or six years, GE has made a very public attempt to cast itself in the role of industrial disruptor, a strategy—sometimes unpopular with investors—closely tied to investments in software (its Predix initiative), sensors and the Internet of things or, as GE likes to call it, the Industrial Internet. Last fall, for example, the company created a $1 billion business devoted to energy efficiency, which Comstock oversees.

Get Data Sheet, Fortune’s technology newsletter.

Earning buy-in required a cultural reset, she acknowledged. “People are fearful.” It also required very specific changes to organizational structures, sales methods, and compensation plans.

Selling GE’s predictive maintenance and equipment optimization services, for example, requires consulting skills that haven’t traditionally be required of GE salespeople. That meant hiring new people and making it safe for them to act differently. “You can train, but you also need new DNA,” she said.

To learn why GE pays people differently for digital jobs, watch:

GE also developed different financial incentives to inspire interest in—and accountability for—its Industrial Internet initiatives, emulating the bonus plans often used by. For example, why not give those driving entirely new sources of revenue a cut of those sales? You’ll find out quickly who really supports the plan, Comstock said.

“Some people want security and they want the upside,” she said. “It’s a good test of who will thrive. How much are they willing to risk? Who will go the distance?”