Perversely, the desire to avoid risk often magnifies it. Dumping money into big me-too projects with modest upside is a lot more perilous than seeding lots of early-stage ideas that are further out on the fringe. In the age of upheaval, incrementalism is the riskiest bet of all. What’s needed is a radical shift in how we think about experimentation. The goal isn’t simply to reduce the uncertainty around new products or get them to market faster, but to build an organization where everyone is working to extend the boundaries of what’s possible. That’s how an organization buys insurance against irrelevance.

Yet in our experience, few companies appreciate the distinction between project risk and portfolio risk. Each potential experiment gets evaluated on its own merits and is expected to clear a high bar of feasibility. That pretty much ensures the company will never invest in the sort of crazy-ass idea that might actually deliver a thousand-fold return.

Remember, the goal is to test your proposed solutions as efficiently as possible, not build something that’s bomb-proof. Nevertheless, you’ll want to be thoughtful about minimizing risks. A few tips:

1. Keep it simple. Test one or two hypotheses at a time, starting with the most critical.

1. Use volunteers. Don’t compel anyone to take part in your experiment.

1. Make it fun. Think of ways to gamify the experience.

1. Start in your own backyard. That will minimize the number of permissions you need and the risk that someone tells you to stop.

1. Run the new in parallel with the old. Don’t blow up the existing process until you’ve validated the new one.

1. Refine and retest. Create an expectation that this will be the first of many experiments.

1. Stay loyal to the problem. Don’t fall in love with your solution. If it doesn’t pan out, search for other testable hacks.

Often a failed project is a critical step in getting to success. Once or twice a year, at our product meetings, I ask all of our managers to complete a simple form outlining their bets from the last few years, divided into three categories: bets that went well, bets that didn’t go well, and open bets. Then we break up into smaller groups and discuss the items in each category and what we’ve learned from each bet. This exercise reminds everyone that they are expected to implement bold ideas and that, as part of the process, some risks won’t pay off. They see that making bets is not a question of individuals’ successes and failures but rather a learning process that, in total, catapults the business forward. It also helps newer people get used to admitting publicly that they screwed up on a bunch of stuff—as we all do.

2. DON’T MAKE A BIG DEAL ABOUT IT

If you make a big deal about a bet that didn’t work out, you’ll shut down all future risk-taking. People will learn that you preach but don’t practice dispersed decision-making… Reed’s reaction is the only type of leadership response that encourages innovative thinking. When a bet fails, the manager must be careful to express interest in the takeaways but no condemnation. Everyone in that room left with two major messages in mind. First, if you take a bet and it fails, Reed will ask you what you learned. Second, if you try out something big and it doesn’t work out, nobody will scream—and you won’t lose your job.

3. ASK HER TO “SUNSHINE” THE FAILURE

If you make a bet and it fails, it’s important to speak openly and frequently about what happened. If you’re the boss, make it clear you expect all failed bets to be detailed out in the open… It’s critical that your employees are continually hearing about the failed bets of others, so that they are encouraged to take bets (that of course might fail) themselves. You can’t have a culture of innovation if you don’t have this. At Netflix, we try to shine a bright light on every failed bet. We encourage employees to write open memos explaining candidly what happened, followed by a description of the lessons learned.

Because failures consume time and resources, you’re smart to use both judiciously. Failures can also threaten reputations. One way to mitigate the reputational cost of failure is to experiment behind closed doors. If you’ve ever tried on a bold new style of clothing to see if it suits you, you probably did it behind the curtain of a store’s changing area. Similarly, most innovation departments and scientific labs are private, with scientists and product designers trying all sorts of crazy things without an audience.