Fail fast to flourish: Lessons from HP, NVIDIA, and Zoho

In 1938, long before they became celebrated tech pioneers, Bill Hewlett and Dave Packard were two 20-something Stanford graduates with a dream—creating their own electronics company. To realize their dream, they rented a 12x18-foot garage in Palo Alto, California to experiment and build a working device that they hoped to sell.

But none of the devices they put together worked. Days and months rolled by, but failure seemed to be the only constant. Undeterred, they pressed on by learning, adapting, and experimenting. A year later, in 1939, the breakthrough arrived in the form of an audio oscillator. What's more, the duo managed to sell the device to Walt Disney for use in the production of the animated musical Fantasia.

In the decades that followed, the Hewlett-Packard Company (HP) grew into a computing powerhouse and remains a household name well into the 21st century. And the unassuming garage, where they battled setback after setback, now stands honored as the "birthplace of Silicon Valley."

From Silicon Valley pioneers to small business owners, this is the persistent truth: Failure is woven into business, and only the undaunted will endure long enough to succeed.

And the way to remain undaunted is embracing failure as part of a continuous cycle of trial, error, and iteration. This is the essence of what came to be known as the fail fast approach.

The "what" and "why" of failing fast

Failing fast is a principle that encourages businesses to test ideas quickly, gather feedback early, and pivot when necessary. The goal is to identify what doesn’t work before investing significant time and resources.

Tech's greatest innovators, from Hewlett and Packard to Steve Jobs, have always regarded failure to be a vital part of the innovation process. To them, learning what doesn’t work is just as valuable as identifying what does. But why is failing fast particularly important now?

We're living in the age of AI, where software development that used to take weeks or months now takes just days. This acceleration has lowered the barrier to entry and unleashed a flood of new products onto the market. On the face of this heightened competition, both startups and established businesses risk being outpaced by rivals who move faster.

To meet and exceed customer expectations now, businesses have to embrace rapid iteration, learning early what works and what doesn’t, and pivoting decisively.

Fail fast, but also fail smart

Having looked at the "what" and "why" of failing fast, next comes the "how."

How can businesses tell apart failing fast and taking reckless risks? The latter entails wasted time and resources—two things that small businesses and startups will always be in short supply of—and result in the stifling of innovation. And without innovation, businesses can't stand out in a crowded market.

The way forward isn't just failing fast, but failing fast and smart. Sim Sitkin, professor of management at Duke University’s Fuqua School of Business, coined the term intelligent failure. Failures are intelligent, according to him, if they satisfy these five criteria:

• Thoughtful design: Experiments are planned so failures yield clear reasons and learnings.
• Genuine uncertainty: Outcomes aren’t known in advance, making each test a true discovery.
• Controlled scale: Projects are kept small to avoid catastrophic losses.
• Timely management: Failures are addressed immediately to identify lessons.
• Broad relevance: Lessons inform not only the current challenge but also shine a light on other parts of the business.

These criteria act as guardrails for businesses as they venture into uncharted waters to innovate boldly but not recklessly.

Cultural ingredients for failing fast

With this broad overview, it's evident that failing fast can't be the belief of only an individual or team in an organization. Failing fast should be embedded into the organization's whole culture. Only then can the organization as a whole—from leadership to front-line employees—be able to align around rapid iteration, early learning, and swift course correction.

But how do you realize this possibility? For this, we turn to a talk by Jensen Huang, co-founder and CEO of NVIDIA, at Stanford University in 2011—a decade before he became one of the faces of the AI boom.

Responding to an attendee's question about the culture at NVIDIA, he points out that nurturing a spirit of innovation across an organization isn't possible unless there's a culture of risk-taking. As he puts it, "I want you to try things even though it is impossible to calculate precisely that it would lead to success."

Huang identifies three prerequisites to foster this culture of risk-taking, which can be regarded as cultural ingredients for every organization trying to fail fast the right way:

• Tolerance for failure: Huang notes that a high tolerance for failure is essential for businesses to fail quickly and change courses when they hit a dead end.
• Intellectual honesty: This requires assessing your business offerings, processes, and strategies on a continuous basis and making necessary changes.
• Teaching risk-taking: Huang emphasizes that risk-taking is both a "matter of skill and a matter of courage." Employees must be taught to take calculated risks, which he describes as failing "quickly and inexpensively."

Related reading:Why the joy of tinkering is important

To see these three cultural ingredients at work, read this blog post by Sridhar Vembu, Zoho's co-founder and chief scientist, from 2012. In it, he announces that the organization will discontinue six applications to "focus our energy and delight our customers with our other products."

On the rationale behind sunsetting these services, he writes: "We are of course, a bit sad to see these products go. But we are a risk-taking company and we like to try many things. We know some of them will not be as successful as we want. We will move the resources (people and otherwise) we had dedicated to these services to our other more promising areas."

This captures the essence of Zoho's convictions. Going against the Silicon Valley grain, we spend twice as much on research and development as we do on sales and marketing, precisely because we've developed a high tolerance for failure. When we encounter instances where products have to be completely rethought or scrapped, we have the intellectual honesty to learn from what didn't work and chart a new course.

We fail fast and fail responsibly while developing new products. Over the years, this approach has helped the organization save money, helped employees grow their skill sets, and helped our customers get more value out of our products. It's a win for everyone.

To wrap things up, Huang offers the final word on the mindset that businesses need to fail fast and innovate:

"Innovation requires a little bit of experimentation. Experimentation requires exploration. Exploration will result in failure. Unless you have a tolerance for failure, you would never experiment. And if you don't ever experiment, you would never innovate. If you don't innovate, you don't succeed. You'll just be a dweeb."

— Jensen Huang, 2011 talk at Stanford University