Steve Blank’s UCSF Class Tests “Evidence-Based Entrepreneurship”
Conventional wisdom says the techniques that help today’s technology startups get to market fast—or, just as likely, fail fast, freeing the founders to start over—don’t work in the life sciences. Rapid product iteration based on agile software development methods; continuous business-model adjustments using customer feedback; cloud outsourcing to increase capital efficiency—all these strategies are seen as somehow inapplicable to the life sciences and healthcare.
Which means companies developing drugs, medical devices, or diagnostic tools are stuck with a slow, expensive, hyper-regulated, insanely high-risk development process. And from this perspective, it’s no wonder that venture investors are getting out of the business, and that first-round funding for life sciences startups is increasingly scarce.
But what if the conventional wisdom is wrong? What if the real problem is much simpler: that life sciences entrepreneurs haven’t been taught how to connect their lab-bench breakthroughs to real customer needs?
That’s the big hypothesis behind Lean Launchpad for Life Sciences & Healthcare, a first-of-its-kind course underway this fall at the University of California, San Francisco. It’s built around the Lean LaunchPad curriculum originally developed by startup guru Steve Blank to help teams of university-based entrepreneurs get their hardware and software ideas to market. But this time around, it’s customized to help teams gain experience turning ideas into companies in the areas of therapeutics, diagnostics, medical devices, and digital health.
The course, which includes teams from several area universities as well as local companies, got underway October 1. The 28 teams admitted to the course met for the second time yesterday at Genentech Hall on the UCSF Mission Bay campus in San Francisco. Competition to get into the course was fierce, and one Arizona-based team drives 12 hours each way to attend class, Blank says.
Blank is directing the course, with teaching and mentoring help from veteran entrepreneurs and venture capitalists such as Abhas Gupta of Mohr Davidow Ventures; Allan May of Emergent Medical Partners; Karl Handelsman of Codon Capital; Jim Hornthal of Hornthal Investment Partners; Todd Morrill of Venture Management Group; Stephanie Marrus, director of the UCSF Entrepeneurship Center; and Jerome Lester, founding director of the Lester Center for Entrepreneurship at UC Berkeley.
The funny thing about the new course is that until pretty recently, Blank himself didn’t believe the startup-building methodology he’d developed—which goes by the loose moniker “customer development”—would work for biotech or medical device companies.
“One thing I say in the The Startup Owner’s Manual is that customer development works for everything except the life sciences,” Blank says. The book, which Blank co-wrote with Bob Dorf in 2012, is a hands-on guide to taking a technical innovation and figuring out whether anyone genuinely wants it. What it actually says is that there are two kinds of early-stage startups: those built around “customer/market risk,” where the question is whether customers will adopt the product, and those facing “invention risk,” where the question is whether the product will even work. Blank and Dorf aimed the advice in the book at companies taking customer/market risk, rather than companies dealing with invention risk, “where it may take five or even 10 years to get a product out of the lab and into production (e.g. biotech).”
But Blank says that in recent conversations with actual life-sciences entrepreneurs and investors, he’s come to think that was an oversight. The UCSF course represents a bet that the customer-development methodology can help companies developing drugs, medical devices, or diagnostic devices evaluate their own markets more realistically, so that they can make better decisions early on about whether and how to proceed.
Essentially, Blank is saying that if early-stage life sciences companies don’t tackle questions about customer/market risk, they may never get a chance to answer the questions about invention risk—or by the time they do, years down the road, they may find that there’s no business case for their product.
“It turns out that the issues are almost the same,” Blank says. “Eventually, it’s about the science—like in semiconductors, where eventually it’s about the process technology or the chip architecture—but if you don’t have a customer, it doesn’t matter. Just as in everything else, entrepreneurs and VCs are wasting enormous amounts of time and money and expertise pursuing the wrong things.”
What’s the solution? Blank calls it “evidence-based entrepreneurship,” a riff on the idea of evidence-based medicine, where epidemiological data informs clinical decision-making.
In technology transfer circles, there’s an understanding that an idea needs to reach a certain “technology readiness level” before entrepreneurs try to take it to market. But it should also have to reach an “investment readiness level,” Blank says. And the process of gauging investment readiness isn’t all that complex or mysterious. It mainly means “getting out of the building,” to use Blank’s signature phrase—talking to actual customers and gathering data. “Do you have a business model? Do you have product-market fit? Do you have validated early orders? Partners? FDA approval? Now, all of a sudden, you have a proper series of metrics.”
Of course, getting to investment readiness is only part of the challenge, especially for biotech startups. Even if it looks like you have a great product, an interested market, and reliable partners, the complexity and distorted incentives of the healthcare industry can blindside you, as Xconomy’s national life sciences editor Luke Timmerman detailed recently in a story about Bexxar, a failed cancer drug that started out as a tantalizing science concept. But some percentage of such failures might be prevented if biotech startup founders took more time early on to study the market, Blank argues.
Blank and his fellow instructors have developed an elaborate syllabus for the 10-week course, which follows the flipped-classroom model: students watch the lecture material online from home, and spend class time discussing the lecture material and presenting the customer feedback they’ve gathered about their startup ideas. (Anyone can watch the lectures, which are drawn from an eight-part video series on Udacity called How to Build a Startup.)
The 28 teams—roughly equally divided between therapeutics, diagnostics, medical devices, and digital health cohorts—are expected to use the time between classes to develop a “minimum viable product” such as a device prototype or a mobile app or website. Each week, they’re required to conduct 10 to 15 in-person or online video interviews with customers. (That’s the “get out of the building” part.) They use what they learn to fill out and update their business model canvas, a graphical tool for visualizing all the internal and external elements needed to build a viable company. In class, they split into cohorts for advanced lectures, weekly presentations, and critiques from instructors and fellow students.
It’s roughly the same structure Blank and other instructors have been using since 2011 for the Innovation Corps, a National Science Foundation program designed to help more NSF grantees get their ideas to market. The Innovation Corps, in turn, grew out of the Lean LaunchPad course Blank co-taught at Stanford University with Jon Feiber of Mohr Davidow Ventures and Ann Miura-Ko of Floodgate in 2010.
In a blog post, Blank writes that what finally convinced him of the need for a dedicated Lean LaunchPad course for life sciences companies was watching a few biotech, device, and digital-health companies go through the Innovation Corps program and successfully navigate major pivots—-changes to their business model canvas or even their core technology. A big pivot can end in catastrophe if a startup waits until after it’s already raised and spent tens of millions of dollars. But an early course change can save years of wasted effort.
Blank admits that digital-health companies and medical-device are probably more agile when the time comes to pivot. “Therapeutics and diagnostics are not on the same level because they’re science-based,” he says. “But on the other hand, what you can learn in 10 weeks is whether the damn thing is worth doing at all. Just teaching PI’s [principal investigators] how to think about partners and targets and markets would be a major accomplishment.”
More than 65 teams applied to join the UCSF course, and the 28 teams selected include 115 students overall. “There are just some amazing teams,” Blank says. “One guy showed up to the interview in surgical scrubs. Turned out he’s the head of surgery at UCSF.”
Blank has briefed officials at the National Institutes of Health about the course. If the teams do well, it could lead to discussions about “franchising” the UCSF course at a number of campuses with heavy NIH funding, the same way the National Science Foundation helped to scale up the original Stanford course.
“Our goal is to get them to start thinking about adopting this as a standard for how you teach commercialization,” Blank says. “For devices and digital-health startups, this is a no-brainer. We’ll see how much benefit we can get out of it for therapeutics and diagnostics—the question is how much getting out of the building is going to help there. But the NIH already believes in at least 50 percent of what we’re doing.”
Already, the reception to the class from student teams and venture investors has been “amazing,” Blank says. “Even more amazing, to be honest, is what an unfilled need this seems to fill. I was the last guy on earth who got it—I just had to show up.”
Here’s a great video interview on customer development in the life sciences between Steve Blank and Iodine co-founder Thomas Goetz, former executive editor of Wired. The interview was part of Rock Health’s Innovation Summit on August 8-9, 2013.
UPDATE 10/11/13: And here’s a video clip from the Oct. 8 session of the UCSF course.