Why Fail Fast Innovation Advice is Wrong

The most popular piece of innovation advice in Silicon Valley is wrong—and it's killing great ideas before they have a chance to succeed.

I can prove it with a story about a glass of water that sat perfectly still while a car bounced beneath it.

My name is Phil McKinney. I spent decades as HP's CTO making billion-dollar innovation decisions, and I learned the hard way that following "fail fast" advice cost us billions and robbed the world of breakthrough technologies.

Today, I'm going to share five specific signs that indicate when an idea deserves patience instead of being killed prematurely. Miss these signs, and you'll become another "fail fast" casualty.

The Water Glass That Changed Everything

So there I was around 2006, sitting in Dr. Bose's lab at Bose Corporation, and he was showing me what honestly looked like just a regular car seat mounted on some automotive hardware. I'm thinking, "Okay, what's the big deal here?"

But then he activates the system and has his assistant start driving over these increasingly aggressive road obstacles. And here's what blew my mind—the car chassis is bouncing around like crazy, but the seat? Perfectly still.

Then Dr. Bose does something that I'll never forget. He places a full glass of water on the seat and tells his assistant to hit a speed bump at thirty miles per hour. The chassis lurches violently, but not a single drop of water spills.

And here's what should terrify every "fail fast" advocate—this technology took fifty years to develop. Dr. Bose began developing the mathematical model in the 1960s. Under today's quarterly Wall Street pressure, this project would have been killed a hundred times over.

When I asked Dr. Bose how he could invest in an idea for fifty years, he explained that keeping Bose private meant they weren't subject to the quarterly results pressure that often destroys patient innovation at public companies.

At HP, we were trapped in that system—and it cost HP billions.

How "Fail Fast" Destroyed Billions at HP

As a public company, we lived and died by quarterly earnings calls. Every ninety days, we had to show growth, and that quarterly drumbeat made us masters at killing promising ideas the moment they didn't produce immediate results.

Let me give you three examples that still keep me up at night:

WebOS: We acquired Palm for one-point-two billion dollars in 2010. Revolutionary interface, years ahead of its time. Killed it when it didn't achieve immediate dominance. Every time you swipe between apps today, you're using thinking we threw away.

Digital cameras: We literally invented the future of photography. Abandoned it the moment smartphones started incorporating cameras.

HP Halo: Immersive telepresence rooms with extraordinary meeting experiences. Sold to Polycom for eighty-nine million in twenty-eleven when quarterly pressures demanded focus. We bought Poly back for three-point-three billion in twenty-twenty-two. We paid thirty-seven times more to reacquire capabilities we built.

We weren't bad managers. We were trapped by the quarterly earnings system that makes "fail fast" the only option for public companies. And it was systematically destroying our breakthrough potential.

Visit Studio Notes over on Substack where I discuss how these quarterly pressures shaped our boardroom decisions and what we were really thinking.

Now, after making these billion-dollar mistakes, I had to figure out how to distinguish between ideas worth killing and ideas worth protecting. What I discovered changed everything—and it comes down to five things I now look for.

When I see all five, I know we've got something worth being patient with. Miss even one, and you're probably wasting your time.

The Five Things I Now Look For

First: Does the Math Actually Work?

Here's how to validate the science without being a scientist yourself.

Start with peer review. Has this been published in reputable journals? Are other researchers building on it? Red flag: if the only validation comes from the inventors themselves.

Next, bring in independent experts. Not consultants who'll tell you what you want to hear—find researchers who have no financial stake in your project. Share your core assumptions with them and ask them to identify any holes.

Look for mathematical elegance. Dr. Bose's suspension model was beautiful in its simplicity. Overly complex models with dozens of variables often hide fundamental flaws.

Here's your action step: Before investing serious mon