Most Science Startups Fail. Here’s Why:

We need to get a lot better at bridging that gap between discovery and commercialization

Share on
BY Greg Satell - 05 May 2018

PHOTO CREDIT: Getty Images

It seems like every day we see or hear about a breakthrough new discovery that will change everything. Some, like perovskites in solar cells and CRISPR are improvements on existing technologies. Others, like quantum computing and graphene promise to open up new horizons encompassing many applications.

Nevertheless, we are still waiting for a true market impact. Quantum computing and graphene have been around for decades and still haven't hit on their "killer app." Perovskite solar cells and CRISPR are newer, but haven't really impacted their industries yet. And those are just the most prominent examples.

The problem isn't necessarily with the discoveries themselves, many of which are truly pathbreaking, but that there's a fundamental difference between discovering an important new phenomenon in the lab and creating value in the marketplace. We need to get a lot better at bridging that gap. To do so, we need to create a new innovation ecosystem for commercializing science.

The Valley Of Death And The Human Problem

The gap between discovery and commercialization is so notorious and fraught with danger that it's been unaffectionately called the "Valley of Death." Part of the problem is that you can't really commercialize a discovery, you can only commercialize a product and those are two very different things.

The truth is that innovation is never a single event, but a process of discovery, engineering and transformation. After something like graphene is discovered in the lab, it needs to be engineered into a useful product and then it has to gain adoption by winning customers in the marketplace. Those three things almost never happen in the same place.

So to bring an important discovery to market, you first need to identify a real world problem it can solve and connect to engineers who can transform it into a viable product or service. Then you need to find customers who are willing to drop whatever else they've been doing and adopt it on a large scale. That takes time, usually about 30 years.

The reason it takes so long is that there is a long list of problems to solve. To create a successful business based on a scientific discovery, you need to get scientists to collaborate effectively with engineers and a host of specialists in other areas, such as manufacturing, distribution and marketing. Those aren't just technology problems, those are human problems. Being able to collaborate effectively is often the most important competitive advantage.

Wrong Industry, Wrong Application

One of the most effective programs for helping to bring discoveries out of the lab is I-Corps. First established by the National Science Foundation (NSF) to help recipients of SBIR grants identify business models for scientific discoveries, it has been such an extraordinary success that the US Congress has mandated its expansion across the federal government.

Based on Steve Blank's lean startup methodology, the program aims to transform scientists into entrepreneurs. It begins with a presentation session, in which each team explains the nature of their discovery and its commercial potential. It's exciting stuff, pathbreaking science with real potential to truly change the world.

The thing is, they invariably get it wrong. Despite their years of work to discover something of significance and their further efforts to apply and receive commercialization grants from the federal government, they fail to come up with a viable application in an industry that wants what they have to offer.

Ironically, much of the success of the I-Corps program is due to these early sessions. Once they realize that they are on the wrong track, they embark on a crash course of customer discovery, interviewing dozens -- and sometimes hundreds -- of customers in search of a business model that actually has a chance of succeeding.

What's startling about the program is that, without it, scientists with important discoveries often wasted years trying to make a business work that never really had a chance in the first place.

The Silicon Valley Myth

Much of the success of Silicon Valley has been based on venture-funded entrepreneurship. Startups with an idea to change the world create an early stage version of the product they want to launch, show it to investors and get funding to bring it to market. Just about every significant tech company was started this way.

Yet most of the success of Silicon Valley has been based on companies that sell either software or consumer gadgets, which are relatively cheap and easy to rapidly prototype. Many scientific startups, however, do not fit into this category. Often, they need millions of dollars to build a prototype and then have to sell to industrial companies with long lead times.

The myth of Silicon Valley is that venture-funded entrepreneurship is a generalizable model that can be applied to every type of business. It is not. In fact, it is a specific model that was conceived in a specific place at a specific time to fund mature technologies for specific markets. It's not a solution that fits every problem.

The truth is that venture funds are very adept with assessing market risk, but not so good at taking on technology risk, especially in hard sciences. That simply isn't what they were set up to do.

We Need A New Innovation Ecosystem For Science Entrepreneurship

In 1945, Vannevar Bush delivered a report, Science, The Endless Frontier, to President Truman, in which he made the persuasive argument that expanding the nation's scientific capacity will expand its economic capacity and well being. His call led, ultimately, to building America's scientific infrastructure, including programs like the NSF and the National Institutes of Health (NIH).

It was Bush's vision that made America a technological superpower. Grants from federal agencies to scientists enabled them to discover new knowledge. Then established businesses and, later, venture backed entrepreneurs would then take those discoveries to bring new products and services to market.

Look at any industry today and its most important technologies were largely shaped by investment from the federal government. Today, however, the challenges are evolving. We're entering a new era of innovation in which technologies like genomics, nanotechnology and robotics are going to reshape traditional industries like energy, healthcare and manufacturing.

That's exciting, but also poses new challenges, because these technologies are ill-suited to the Silicon Valley model of venture-funded entrepreneurship and need help to them get past the Valley of Death. So we need to build a new innovation ecosystem on top of the scientific architecture Bush created for the post-war world.

There have been encouraging signs. New programs like I-Corps, the Manufacturing Institutes, Cyclotron Road and Chain Reaction are beginning to help fill the gap. Still much more needs to be done, especially at the state and local level to help build regional hubs for specific industries, if we are going to be nearly as successful in the 21st century as were were in the 20th.