Getting Failure Right
The best companies learn from failure. That's not a fluke.
Is there a right way to fail? Successful companies would say yes.
Take Amazon for an example…
In 1999 Amazon was on. The company revenue grew 169% to $1.64 billion. But the company was not satisfied with its platform.
eBay, the rising auction platform was dominating the world of ecommerce, and capturing the eye of investors. Although eBay’s it was just 1/7th of Amazon’s revenue, its market cap was over $16 billion, compared to Amazon’s at $20 billion. Clearly investors valued eBay’s model and market.
Amazon saw auctions as a growth opportunity and quickly launched its own auction service, Amazon Auctions. The strategy was simple: leverage Amazon's growing customer base and brand recognition to challenge eBay's stronghold.
However, despite Amazon's strengths, the auction platform failed to gain significant traction and was ultimately discontinued after a relatively short lifespan. By every outside measure, Amazon Auctions was a failure.
But the story didn’t end there for Amazon.
The company’s foray into the auction world proved to be a valuable learning experience for Amazon. From its experience in auctions, Amazon learned key insights into managing third party sells, building trust and community, using data insights that led to building Amazon Marketplace, its platform for third-party vendors selling on Amazon’s platform.
Today, Amazon Marketplace makes up 60% of Amazon’s sales. That’s over $385 billion in 2024!
The ability be resilient to and learn from failure has been critical for Amazon, whose failures have been wide-ranging. In many of those instances, the failure of one product (example, Amazon’s Fire Phone), has been directly linked to success in other products (Amazon successfully integrated its technology into the Echo smart speaker series, which has been highly successful).
Amazon is not alone in its ability to turn failures into success. Apple’s first handheld product, the Newton (1993) was a commercial failure but eventually Apple launched the iPod, iPhone, and iPad, producing hundreds of billions in revenue. Microsoft has always been known as a company that turns every product failure, or two, into a success.
Failure has been a consistent part of the history of many companies, especially startups. Take for example, two popular ridesharing companies that initially missed the mark on their original product:
Lyft started out as a service of Zimride, a long-distance intercity carpooling company founded in 2007 by Logan Green and John Zimmer. Zimride focused on providing ride-sharing services between college campuses, inspired by Green's experiences using informal ride-sharing networks during his college days and his observations of similar systems in Zimbabwe.
Uber started as a service for calling high-end “black car” services, it wasn’t until the platform was running that its team found a wider application.
The approach of these and other companies show that failure can be a productive tool in innovation. The problem with failure isn't failure itself. The problem is not using failure to your advantage. You will learn how to flip failure from a major problem to avoid into your greatest teacher.
Facing facts:
First, no one likes to fail. In fact, we inherently hate it.
Second, because of that we tend to cover it up. This causes bigger problems, like, at a minimum, we don’t learn adequately from our failures.
Third, there are underlying psychological factors that play heavily into our future success.
For example there’s the Streisand Effect (which we’ll talk about in the future), which shows that when we cover up or fail to learn from our failures, we end up spending more time focusing on them, which in turn governs future behavior.
And then there’s Attribution Theory, which tells us how poorly people explain their successes and failures. It's particularly relevant because individuals who attribute failures to specific, changeable factors (rather than permanent, personal characteristics) are more likely to learn and improve from their mistakes. This impedes our ability to properly analyze failures more constructively.
Founders tend to have particularly distorted views of risk and failure. We use bad analogies like jumping off cliffs to build airplanes which unnecessarily promote reckless behavior with other people’s money (a bad distortion of the agency effect).
This is often promoted by the venture capital community, which has not a different risk profile but a different way of harnessing risk than founders.
The result is that we don’t tend to understand how to fail right.
So, in the next few articles we are going to examine failure more generally and how to turn our view towards learning, quickly, cheaply, and painlessly. We’ll cover in particular how to:
Create the space for psychological safety. When team members fear judgment or repercussions, they hide mistakes and avoid taking necessary risks. Creating psychological safety means actively demonstrating that vulnerability is welcomed, and mistakes are viewed as learning opportunities. Leaders in particular should model this behavior by openly discussing their own failures and the lessons learned, showing that candor and growth matter more than perfection.
Get rid of fail fast — and start reframing to learning fast. “Fail fast” is a snappy slogan for a sloppy strategy. Rather than pushing teams to “fail fast,” which can trigger stress responses and poor decision-making, focus on “learning fast.” This means starting with small, low-cost experiments that provide quick feedback. By reframing the goal as rapid learning rather than failure, teams stay cognitively engaged and creative rather than defensive and risk averse. Use true minimal viable products (MVPs) to gather insights efficiently without the pressure of perfection.
Create a learning environment. A true learning environment goes beyond just tolerating mistakes — it actively extracts value from them. Implement regular retrospectives, blameless post-mortems, and knowledge-sharing sessions. Make learning a deliberate practice by setting aside dedicated time for teams to reflect, document insights, and adapt processes based on what they discover. Celebrate instances where lessons learned led to improved outcomes.
Build advantages through speed and observation. Success comes not from avoiding all mistakes, but from detecting and responding to them quickly. Develop strong observation systems that catch issues early, when they're small and manageable. Train teams to recognize warning signs and patterns. The faster you can observe, analyze, and adapt, the more competitive advantage you build. This creates a virtuous cycle where rapid learning leads to better decision-making.
Building the right kind of risk-taking model. Use the barbell strategy to balance risk: keep most activities in the safe, proven zone while dedicating a smaller portion to controlled experiments and innovation. This approach protects core operations while still enabling breakthrough discoveries. Clearly define which initiatives fall into which category, so teams understand when to be conservative versus when they have permission to push boundaries and take calculated risks.
I hope you’ll join me on this journey, and even share it with others. To be honest, this series could be a disaster, but I am willing to take that risk and learn from it.
To start, here are a few questions for you to share with your team and start discussions.
Things to consider:
How does your team respond when we experience a failure or something goes wrong?
Do our people feel safe to admit mistakes?
What formal processes exist to learn from failures?
Are lessons actually implemented, or does our organization repeat the same mistakes?
Until next time, lead with purpose.
Will
About Leading Matters:
Leading Matters is the trusted source for aspiring and seasoned leaders alike, providing them with the tools, insights, and inspiration to become intentional leaders that build more innovative, engaging, and agile organizations.
#innovation#transformation#founders#CEOs #failure #leadingmatters #failfast #learning #biases #leaders #leading