There has been so much written on "failing fast" as a way to operate a startup, the phrase has officially become cliche. Silicon Valley deserves at least some of the blame (okay, a lot of it) for rendering this concept nearly meaningless.
This isn't the first time this has happened with an otherwise compelling idea. When it does, though, I know it's a good opportunity to return to the source of the idea and re-extract its usefulness.
I recently did this at my company, Virta Health. If failing--and what I really mean in this context is experimentation and applied learning--is so critical, I suspected I should see a bunch of discarded ideas alongside our progress.
I ended up seeing fewer "bad" ideas than I expected, so to encourage failure, I offered cold, hard cash for the best examples of trying a new initiative, an unconventional idea, or process.
What resulted was not only faster iteration, but also a stark reminder of three important principles to leverage failure for positive growth.
Three ways to use failure to grow:
1. Build it into your core values.
Most of the time your starting point should be the knowledge gained by others before you. With that said, in many cases, especially if you are creating a new market, you won't always have a roadmap to guide decisions and execution.
It may take 10 tries to arrive at the right answer. This means that if an employee has a big decision to make and brings a solution with zero testing, red flags should go up.
To address this, your company needs to weave failure into your core values. At my company, we call it "risk-taking and rapid iteration." However you decide to name it, experimentation should be part of what makes something complete, as it's a massive driver of better decision-making and outcomes.
2. Train your leaders.
To successfully scale, team leaders have to delegate nearly 100% of an initiative. In this situation, even with proper coaching, people will make mistakes. To be successful at experimentation, employees need the psychological safety to know that perceived failures are a critical part of the growth process.
Don't assume that people innately have it in them to be comfortable with seeing several ideas fail, either as a manager or the executor of these ideas. Your company should train team leads on how to encourage and evaluate experimentation. Doing so helps your employees, and your company, grow healthily.
3. Use failure as an exclusion lens.
Products naturally tend toward increasing complexity, unless focused discipline is applied to continually prune unnecessary features. In addition to using your experimentation framework to figure out what to add, use it as an exclusion lens. What's not working or necessary, and how do you make your product and operations ever more simple in order to continually delight and serve your users?
The truth is, just about everything you do at your company can be made better through a well-thought out "failure framework." With that said, there are two types of failure that should be avoided at all costs.
Two types of failure to avoid at all costs
1. Slow failure
Taking months to realize a mistake--either because of ignorance, slow execution, or poor follow-up--that could have been tested in hours or days, can be the death of a company. Design your tests for speed, execute them, make a decision, and move on.
Beyond the practical advantages of this approach, it's also emotionally easier, as you're not stuck dwelling on decisions that could have been made in a matter of hours with a simple iteration playbook.
2. Critical failure
It should go without saying that something that can sink an entire company should not be subject to crazy risk taking. As an example, at my company, one obvious place we'll never take risks is with patient health.
If it wasn't apparent by now, offering cash for failure was a way of testing my own advice. When I announced it, I admitted it might be a bad idea, but no matter what, I'd learn something valuable about what mattered for our patients' and employees' happiness.
At least on this latter point, which is the one that matters, I was right.