While competition is heating up—with players like Toyota, General Motors, Alphabet, and Tesla setting ambitious goals and making big bets—the question remains: Are Americans ready for driverless cars? Will we ever be?

As I’ve studied complex organizational transformations over the past decade, I’ve come to recognize what must happen to create the behavioral change that makes adoption of new technologies successful. A societal shift toward self-driving vehicles will require such massive behavioral change, especially as trust continues to plummet. But just because we wouldn’t get behind that self-turning wheel today, doesn’t mean we wouldn’t take that chance tomorrow.

Here are two things that need to happen to make self-driving cars a reality:

Trust must be earned

With the adoption of any new technology, establishing trust with end users is table stakes. And just as Rome wasn’t built in a day, trust isn’t established overnight.

Luckily for self-driving cars, several important milestones have already been reached on the road to fostering consumer trust in autonomous vehicles—notably, our reliance on GPS and the growing use of ride services like Uber and Lyft.

Riders learned to trust ride services because fears were mitigated. GPS functions ensured a high rate of ride success; transparency of location created a sense of safety for riders in the presence of a complete stranger; and changes in pickup and dropoff locations were seamless. In other words, many of the questions underlying Uber’s technology had been answered before the first rider hailed a car.

The same story holds true for self-driving cars. Like climbing a ladder, driverless technology will have to be proven to people step by step before the ultimate rollout. Drivers aren’t going to take their hands off the wheel without a solid foundation of trust.

American automobile consumers have already learned to trust technology such as vehicle backup cameras, lane departure assists, and blind spot detection, all of which assist with building a new relationship between car and driver. Few of us realize that automobile manufacturers have been prying our hands from the steering wheel one useful piece of technology at a time.

Still, it will be difficult to make consumers comfortable with autonomous vehicles if lingering fears remain about the safety and reliability of the underlying technology. In order to have confidence, consumers need to see a small group of engaged, early adopters demonstrate the technology’s usefulness and safety, and communicate it to the masses.

While technology innovators, who have been first movers when it comes to autonomous vehicles, are typically more comfortable with the risks associated with cutting-edge technology, these early adopters are selective about which new technologies they invest in. Continued negative news about the safety of self-driving cars might keep them away—and this early adopter pool, according to Everett Rogers’s model, is essential for building the communication networks necessary to roll out innovations more broadly throughout society.

A perfect storm

Seventy-five million people now use Uber each month. What enabled this widespread adoption? Was it driven by intention—or chance?

The answer is: a little bit of both. Uber didn’t single-handedly enable the behavioral change that allowed its technology to take off. It entered the market when mass transit was already beginning to fail, urbanization was increasing, and demand for public transit was through the roof. It was this perfect storm of events that fueled the company’s success.

While the tech-enablement of any task is now accepted as par for the course, for self-driving cars to truly become a reality, we’ll need that same kind of perfect storm. Increasing demands on people’s time, greater automation of everyday life, advancements in driver-assistive technologies, and continued urbanization will create the urgency needed for the rollout of driverless technology and adoption of new driver behaviors.

But first, leaders in the field must understand the process through which consumers might embrace self-driving cars: trust in the technology, confidence in its use, and the convergence of societal and environmental pressures. All of these factors must come together to drive the large-scale behavior change necessary for driverless cars to go mainstream.

Eric Ellis is a principal at Kotter, a leadership and strategy acceleration firm.