Why (and how) the smartest leaders encourage failure

“If the size of your failures isn’t growing you’re not going to be inventing at a size that can actually move the needle.” 

Jeff Bezos’s words—written in a 2019 letter to shareholders—suggest a more clear-eyed view of the innovation process than the paradoxical perspectives of many other senior executives. 

Oh sure, CEOs agree that innovation is important. In fact, 92% say it’s a top priority, according to a recent McKinsey article. But at the same time, more than 90% of CEOs say they do a lousy job at innovation. The reason for this confusing response can be boiled down to one major point, alluded to by Bezos: 

Fear of failure.

Yes, fear of failure—and wariness of the mixed messages they get from management. You can’t expect people to take risks, challenge the status quo, and explore new ways of doing things when you measure them on hitting near-term targets with near-perfect accuracy. Innovation requires curiosity, experimentation, and learning—the trifecta I call, “try, fail, learn.” Inevitably, projects will fail; people will fail, too.  It’s normal, and it’s high time we normalized it in business. 

Below are five ways you can put meaningful metrics in place to incentivize healthy risk-taking and smart failure in your organization. 

1. Start Small: Create Rituals That Normalize Failure

Changing culture starts with small, visible experiments that make failure feel safe, expected, and even energizing. One of the simplest and most effective practices I’ve implemented is what I call “Fail-Free Fridays.”

These are dedicated 60-minute blocks of time where teams meet weekly to talk about what’s not working and share ideas about things they want to try. No PowerPoints. No success criteria. No approvals. The goal isn’t to solve the problems or produce a breakthrough; it’s to openly discuss what’s not going well and experiment with new ideas. Without fear.

How to make it measurable:

• Track the number of problems discussed
• Track the number of ideas generated
• Track self-reported psychological safety (before and after)
• Track cross-functional collaborations initiated during these sessions

2. Define What a ‘Good Failure’ Looks Like

Not all failure is equal: Experimental failure is necessary for learning and invention, whereas operational failure is due to poor execution, lack of discipline, or not following processes and procedures. Help your team by painting a picture of what “good” failure looks like. Find a recent example and do a post-mortem analysis by showing how the initiative:

1. Was aligned with strategic priorities
2. Was based on a clear hypothesis
3. Was a controlled experiment with defined parameters
4. Produced a documented learning
5. Informed future decisions

The next step is to measure the proportion of failures that meet these criteria.

Sample metrics might include:

• % of failed projects with clear hypotheses
• % of failed projects that produced specific, documented learnings
• Estimated resource savings from ideas invalidated early
• Time saved by early “no-go” decisions compared to traditional project lifecycles

3. Reward Learning Behaviors, Not Just Outcomes

Traditional performance reviews reward outcomes: sales targets met, product launches delivered, efficiency increased. These metrics reinforce predictability—which is essential for operations but corrosive to innovation.

To incentivize smart failure, organizations must introduce behavior-based performance metrics tied to learning and experimentation.

Examples include:

• Number of experiments initiated or proposed
• Willingness to challenge outdated assumptions or raise contrarian ideas
• Speed of testing a new idea—how quickly a team can test, learn, and adapt
• Cross-functional collaboration and knowledge-sharing

One technique I’ve used is integrating a “Learning Objectives” section into performance goals. Employees must identify one or two areas where they will experiment, explore, or test new approaches—and leaders evaluate how intentionally and transparently they learn from the results.

Behavior-based metrics shift attention from “Did you succeed?” to “How did you learn, and what value did that learning create?”

4. Build Transparency Into the System: Share Failures Publicly with Leaders as Role Models

For failure to be normalized, it must be visible and leaders must be role models showing how it leads to learning and growth.

Examples of transparency-building mechanisms:

• Town Hall or All Hands Meetings where the leader dedicates 15 minutes of the agenda to allow an employee to share a story of failure and learning (leaders can share their stories, too)
• Monthly “Lessons Learned Roundtables” where teams briefly share one failed experiment and one insight
• A digital “Failure Dashboard” highlighting experiments run, hypotheses tested, learnings extracted, and next steps
• Internal newsletters profiling teams who tried something bold, failed smart, and moved the organization forward

Metrics here can include:

• Number of learnings shared across business units
• Participation rates in roundtables or learning forums
• Cross-team adoption of insights
• Repeat failure rate (a powerful metric—if it decreases, organizational learning is improving)

5. Make Failure Economically Visible: Track the ROI of Learning

We talk a lot about Return on Investment (ROI) of new projects. Similarly, the most important, and most neglected step is quantifying the Return on Failure (ROF).

Leaders know that invalidating a bad idea quickly is just as valuable as scaling a good idea. In many cases, it’s more valuable. Early failure prevents wasted resources, prevents misaligned investments, and accelerates strategic focus.

Organizations can track:

• Cost savings from early project termination
• Time-to-decision (how fast the organization can rule in or rule out an idea)
• Increase in pipeline throughput (better quality ideas lead to more opportunities making it to market)
• Portfolio health metrics (percentage of projects in exploratory vs. execution mode)

The Cultural Shift: From Fear to Learning and Growth

The goal is not to create a workplace where failure is unbounded or unexamined. The goal is to create a workplace where learning is measured, rewarded, and operationalized.

When failure is treated as data—not deficiency—organizations accelerate innovation, attract bolder thinkers, and build resilience into their strategy. They become more adaptive, more opportunistic, and more capable of navigating uncertainty.

Leaders who want sustained growth don’t ask, “How do we avoid failure?” They ask, “How do we create more opportunities to learn—and how do we measure the value of that learning?”

The takeaways? Start small. Measure early. Reward curiosity. Make learning visible. Treat disciplined failure as a strategic asset.

Organizations that do this consistently don’t just innovate—they grow, consistently and over time. That’s what successful failure can do for your business.

View the full article