Don’t Fear Something Going Wrong

Worry instead about what happens when it inevitably does.

A few years ago, I shared a stage with Michael Lopez-Alegria to talk about risk-taking. Michael is an astronaut. He’s flown three space shuttle missions, spent 215 days on the International Space Station, and logged more than 67 hours of spacewalks. He knows a little bit about taking risks.

But strangely enough, Michael’s approach to risk-taking is completely different than my own.   I’m used to operating in an environment where it’s common to push out products and ideas that may be only partially finished or unproven. I’m comfortable with the idea that most of my initiatives will fail, viewing such failures as opportunities for learning and growth.

Michael, on the other hand, has spent his career working within the confines of the US Space program, where billions of dollars are spent on comprehensive safety measures: extensive testing, redundancy in critical systems, years-long astronaut training, and meticulous mission planning. He’s comfortable seeing launches routinely scrubbed for failures of even relatively trivial back-up systems.

Clearly, space travel involves taking risks, just as entrepreneurship does. But our respective experiences with risk—and our approaches to it—have been so different that it made me wonder if we were talking about the same thing. Since one definition of risk is being willing to do something where you don’t know the outcome in advance, are these two types of risk really the same?

I’ve concluded that they are not. But worse, I believe that confusion between the two types of risk taking is what prevents so many aspiring entrepreneurs from even starting. Too many entrepreneurs believe that – like a space mission – they can’t start until they have eliminated as many unknowns as possible, a process which is so time consuming, so technically difficult, and so expensive – that unless you are independently wealthy, your great idea is never going to get off the launch pad.

When most people think about risk, they are scared that something is going to go wrong.  But what they should be thinking about is, what are the consequences when it inevitably does?

In NASA’s case, space travel requires intricate spacecraft, complex propulsion systems, and myriad interconnected systems, all expected to work in an extremely harsh environment. With so many things that can go wrong, it’s likely something will.

But because the consequences of failure are so high—whether in terms of failed mission objectives, loss of life, or future public scrutiny—NASA simply can’t afford to let something catastrophic happen. They want as much as they can to go as planned, and they’re willing to accept the extraordinary costs and lengthy timetables involved in getting to that level of certainty.

Entrepreneurship is at the opposite end of this continuum. Like a space launch, I fully expect that something is going to go wrong with almost everything I try. But unlike NASA, the consequences of my failures are minimal. Better yet, I’ve learned that even the most abject failures result in some positive learning that informs my future efforts.

It's not a given that certain categories carry more consequential risks than others. One big reason that SpaceX has become the go-to company for space launches is that Elon Musk simply applies the principles of entrepreneurship to space travel. He moves quickly, sets low expectations for each effort, and is perfectly willing to accept the consequences of repeated failure.  (When the latest SpaceX Starship exploded just after stage separation, SpaceX reported they were “pleased” with the flight, despite its “rapid unscheduled disassembly.”)

Self-driving cars have also taken a more progressive approach to risk. Tesla and Waymo, for instance, are willing to put their vehicles on public roads with beta-level driver-assistance systems, gradually progressing toward full autonomy. These companies believe that real-world testing is essential to gather data and improve their technology. They aim to balance the potential risks (even though these risks are potentially severe) with the societal benefits of advancing autonomous driving technology.

In the case of Pharma, the amount of time and resources required for drug development are driven by the social and legal imperatives that we minimize the risk of harm to people. We want to know that a new drug works, safely, before we have people use it. As a result, bringing a new drug to market can take more than 10 years and cost well north of one billion dollars, so drug development is a game only the largest pharmaceutical companies can play.

As I wrote last week, I see the same thing in banking, insurance, and every other highly regulated industry, where we’ve decided as a society (or, more accurately, lawmakers have decided for us) that we need 100% reliability—if only because the consequences of failure are so severe.

In most cases, like highway safety, we set a line somewhere in the middle, mandating that cars include seat belts, child safety seat systems, electronic stability control, and front airbags, but stopping short of trying to eliminate highway fatalities entirely, which although theoretically possible, is prohibitively expensive, potentially costing millions of dollars per mile.

In the end, the level of risk you are willing to accept comes down to three questions: How likely it is that what you are trying is going to fail? What are the consequences if it does? And most importantly, how well can you handle that failure?

But practically speaking, unless you’re in the business of developing new drugs, building new financial products, or devising new ways to send people into space, don’t let yourself get so hung up on building something that “can’t fail” that you never actually start.

Instead of focusing on your “big idea,” think about the ways to minimize the consequences of its inevitable failure so you can keep your powder try for the next try.  Make your tests easy, simple, quick, and cheap. Test them on small samples. Do whatever you can to increase the number of cycles you get.   

It’s that easy.

Or at the very least, as I’m sure Michael Lopez-Alegria would agree, it’s definitely not rocket science.