When tech entrepreneurs attack science

(Cross posted from the Bio Business blog)

One of the most remarkable trends in science-based entrepreneurship is the recent explosion of fake meat companies.

I saw my first fake meat company in 2015 at the graduation event for the first class of startups at the IndieBio accelerator in San Francisco, doing egg whites. Now leading companies like Impossible Foods and Beyond Meat have landed their fake hamburger in fast food chains.

At breakfast yesterday with my boss’s boss, he remarked that some of the burgers are actually quite good, and we agreed that (someday) it has the potential to be a trillion dollar business. This seems to be one of the rare examples where the outrageous predictions by tech entrepreneurs of creating a huge new market might actually be true.

Adoption Paths

The initial pioneers may grab decent exit values, and the long-term future of replacing meat seems compelling. California alone spends 6 trillion liters of water a year on alfalfa alone (feed for cattle and horses), not counting other states, and the water for pig slop and chicken feed. Meanwhile, cow farts play a non-neglgible role in increasing greenhouse gasses. And there is also a sizable niche of the populace that either refused to eat meat, or even wants to deny others the right to do so.

It’s not clear when it will become a trillion dollar market: as I showed in my 2014 paper in the Journal of Technology Transfer, California firms created the solar industry but flamed out because they got into the market 20-30 years too early. Competing with commodity electrons is a tough adoption curve: very few people voluntarily choose to pay 50% or 100% more than market prices for a commodity, although Germany and California show that politicians can force their voters to do so and (mostly) get away with it.

What I didn’t realize until I thought it through is that meat has an easier adoption curve, with a wide range of niche markets that can be sustained at premium prices. You have affluent people who don’t eat meat — or, even better, recently gave up meat — as well as environmentally conscious customers who would like to avoid meat. You have people who are willing to give it a try, out of curiosity. And — as the burger joints have demonstrated — the B2B customer (distribution) is willing to try a niche product to raise average selling prices.

Thus, as the product gets better and the prices get lower, these firms can establish and grow their beachhead in the food market, carving off ever-larger segments of the market. Funded by Sand Hill Road and led by ambitious entrepreneurs, some will hold off for Facebook-style IPOs, but many of the weaker players will be bought up by ADM, ConAgra, Hormel and the like — providing bottomless capital to spur innovation and adoption. (The entry barriers are low enough that Tyson Foods is launching its own product directly, rather than by acquisition).

We Need Science

However, to fully displace meat, there are major technical challenges to be overcome, both in quality and cost. I supervised a student project to research synthetic organs — a more demanding applications — but still getting the texture right will require both science (new insights) and engineering (new applications) to create a quality product at a competitive price.

Thus, I was struck by the decision of one fake meat company to attack GMOs to win market share. Yes, the CEO is a 24-year-old recent Berkeley grad who’s never worked in a company. Yes, her bachelor’s degrees are in toxicology and environmental studies rather than molecular biology or chemical engineering. But the company does have one PhD (food science) in its leadership, so they presumably are doing actual science.

It reminds me (and not in a good way) of the various surveys that showed the gap between what the public thinks and what scientists (writ large) think about GMOs, including a 2015 survey that said 37% of the public thought GMOs are safe vs. 88% of scientists.

More troubling is that the certainty of these opinions seems inversely proportional to actual knowledge. As the NY Times wrote on Monday:

In a paper published early this year in Nature Human Behavior, scientists asked 500 Americans what they thought about foods that contained genetically modified organisms.

The vast majority, more than 90 percent, opposed their use. This belief is in conflict with the consensus of scientists. Almost 90 percent of them believe G.M.O.s are safe — and can be of great benefit.

The second finding of the study was more eye-opening. Those who were most opposed to genetically modified foods believed they were the most knowledgeable about this issue, yet scored the lowest on actual tests of scientific knowledge.

In other words, those with the least understanding of science had the most science-opposed views, but thought they knew the most. Lest anyone think this is only an American phenomenon, the study was also conducted in France and Germany, with similar results.

So I get that trust in authority has been declining since the 1970s. I get that we have many people who don’t understand — or have the time to personally verify — scientific research. And, as Orwell predicted (and Goebbels proved), people are easily persuaded to believe lies if they are repeated often enough in the mass media.

Still, why would companies that depend on scientists to create their products help promote such lies? Isn’t the benefit of saving the planet enough, without having to rely on junk science for the purpose of virtue signaling? And if companies that depend on science attack science, what are the implications for K-12 and university science indication, science policy and the idea of using facts — rather than emotion - as the basis for making science policy?

Incubating, accelerating and funding life science startups

On Thursday I visited the SF Indie Bio Demo Day. The San Francisco life science accelerator was graduating its first class of 12 early-stage startups.

Indie Bio was launched and funded by SOSVentures, a $200m seed fund based in Cork, Ireland. Its launch was seen as a reaction to the decision of Y Combinator to take on 5 biotech firms last summer, although Indie Bio founders insisted that their plans predated Y Combinator’s interests.

Y Combinator
The Mountain View-based Y Combinator is widely considered to have invented the accelerator format — or at least to have perfected it with its 2005 launch in Silicon Valley. Its model includes:

• Cohort admission process (two cohorts per year)
• Incubator-style office space combined with seed-stage equity funding
• Mentoring services to help firms refine their product and strategy
• Mandatory on-site participation of the CEO during the intensive 3-month process
• At the end of the incubation period, a hype-filled “Demo Day” graduation ceremony that both brands and publicizes companies being kicked out of the nest

Most of its 800+ alumni have been software or software-enabled companies that fit its original focus. Its best known graduates are online services companies such as Dropbox, Airbnb, Reddit and Weebly.

While Y Combinator has been a great success — in terms of publicity, investments and picking (and perhaps helping†) winners, there have been questions about whether the model could be used for companies with longer development cycles. There are also obvious economies of scale and scope if each cohort has overlapping technology orientations.

Y Combinator had some successes in hardware — and now a handful of recent life science graduates — but IMHO the generalizability remains an open question.

† There is a longstanding debate in higher ed whether schools such as Stanford and the Ivies mold their students into successes, or merely pick students who are bound to succeed no matter what. There are obvious parallels to angels, VCs and incubators

Indie Bio
As a former software entrepreneur who has spent the past four years teaching business at a biotech graduate school, I was eager to see how the Silicon Valley formula would be adapted for starting life science companies. On average, life science startups are different — with greater technical uncertainty, capital costs, development costs and time to market.

SF Indie Bio has adapted the accelerator format of Y Combinator: in some cases, the language and terms are almost identical. However, a crucial difference is the need for wet labs; healthcare IT companies can be incubated at a software incubator, but therapeutics, device and most diagnostic companies cannot. Designing and staffing a wet lab isn't cheap, and every life science incubator faces the challenges of which equipment is needed and how to cover it.

Another advantage for Indie Bio is an (acclaimed) life science-specific incubator team. Ryan Bethencourt and Ron Shigeta earlier started the Berkeley Biolabs, while Arvind Gupta is a venture partner with the parent venture fund. At this week’s Demo Day, there were clearly synergies between the companies in the first cohort, particularly for the tools companies that had ready-made customers while those customers saved capital as free beta sites.

Ron Shigeta, Ryan Bethencourt and Arvind Gupta at Thursday’s rollout

As with other Bay Area incubators and accelerators, the Indie Bio location means that tenant companies are close to a wide range of potential investors.

The companies in the first round received $100K of cash and in kind funding in exchange for an 8% equity stake. The next round of firms will get this, and also the option of a $150K of cash in exchange for a convertible note.

Seizing a Place in the Value Chain
After visiting Indie Bio — and talking to its founders and entrepreneurs — I got a better sense of where an accelerator would fit in the entrepreneurial value chain of a life science company. (Other life science incubator/accelerators seem to have been launched in Berlin, Houston, Israel, and Winnipeg.)

As the name suggests, an accelerator can make a big difference in accelerated a firm's growth. However, it covers only one brief phase of a longer process of development.

Indie Bio focuses on the steps of the lean startup process where entrepreneurs develop the minimum viable product and decide whether to proceed or pivot. However, before firms are ready to join an accelerate they will need to develop their science. This could be at a university, at a biohacker space (such as BioCurious), or a local life science incubator such as QB3 or J&J's JLabs — or the Berkeley BioLabs cofounded by Bethencourt and Shigeta.

At the same time, once they leave the accelerator they will need a home. If the company is small (< 5 employees), it might still fit in an incubator. If they hire multiple employees, they're too big for either an incubator or accelerator and will need an actual office. Either way — as earlier in their development cycle — they will need a shared wet lab rather than just (as with a software or healthcare IT company) access to cloud servers.

Bethencourt and Shigeta are aware of this imperative. They’re working to identify an affiliate or partner facility that the Indie Bio companies can graduate into.

New models of biotech entrepreneurship

Thursday night, the Oxbridge Biotech Roundtable concluded its 2014 Onestart Americas $150k business plan competition. Several dozen contestants, mentors, and one LA-area biotech entrepreneurship professor travelled to the City Club in downtown San Francisco to hear the 10 finalists offer elevator pitches, and then see the money awarded to one of the teams.

Following the initial (2013) competition in London, the American competition began with December's submission of pre-proposals by 150 teams. 35 of these teams (including one led by current KGI MBS students) travelled to a February bootcamp at Stanford for additional training and mentoring, before the 10 finalists were announced in April. One of the 10 finalists, Excell Biosciences, was cofounded by a KGI alumnus.

As in the two previous Onestart Europe competitions, the contestants were required to be aged 35 or younger. As in Europe, the US competition was co-sponsored by SR One, the corporate investing arm of GlaxoSmithKline.

The 10 US finalists reflected an interesting mix of geographies and industry segments. From the leading U.S. biotech clusters, only three teams were from the Bay Area, one from Boston and none from San Diego. The competition also included two teams from Toronto, one from Vancouver and one each from Los Angeles, Denver and New York.

The mix of products was broadly representative of life science startup companies

• 3 therapeutics
• 3 devices
• 1 diagnostic
• 1 healthcare IT (consumer app)
• 2 process innovations, for manufacturing and for drug delivery

Most of the startups were in some phase of bootstrap funding, and the judges commiserated with the particular difficulty that therapeutic companies face in raising the tens (or hundreds) of millions necessary to come to market.

Have judged, organized and mentored business plan competitions for years, I was struck by several aspects of the OBR finalists compare to typical (college-based) business plan competitions.

• Of course, a lll of the plans were about technology. That’s true for our KGI competition but not for the typical b-school competition.
• Second was the depth of the entrepreneurs’ understanding of their technology. Again, at many (not all) b-school competitions, the entrepreneurs are smart individuals who BS their way through a partly thought out concept. On Thursday, the winning entry — Resilience Pharmaceuticals— reflected six years of work, including the 2013 MIT PhD dissertation by cofounder Retsina Meyer.
• The judges remarked on the depth of the teams (and that, like VCs, ultimately they had to bet on the teams as much as the ideas). The audience only saw one team member make a presentation, but some teams had five or more official members listed in the program. At least four teams were headed by polished PhDs. Apparently the winning team has attracted Boston veterans to join their team, beyond the founders.
• Finally, some firms had won equity investment prior to the finals. The winner has attracted a commitment from Third Rock Ventures, a leading biotech VC.

Representing SR One, partner Matthew Foy said that the $150K prize “is not the point of Onestart — it was just the carrot to get people’s attention”. They seem to have succeeded in doing so. Overall, in its second year (and third competition), the Onestart sponsors seem to have moved closer to their goal of creating actual entrepreneurs and startups.

It will take a few months to see how many of these 10 companies will actually launch and — more importantly — how many can succeed in bringing a product to market. Still, in terms of the structure of the program (and the nature of the competitors), the sponsors seem to have already done better than all but a handful of school-based competitions.

Tech startups: cross-functional people or cross-functional teams?

Today was the culmination of the business plan class (ALS 458) here at KGI, with the final presentations by 6 teams — some of whom will be going on to formal business plan competitions elsewhere. So it was the day of the year that the students, and invited guests most celebrate (and ponder) the nature of tech startups.

Our students are unusual in having both science and business training: they come with a science (or engineering) degree, they take science classes, and they take business classes. So in effect, they are cross-functional individuals, with a little bit of knowledge about a lot of things in their heads. Similarly, a company like HP used to pick their marketing staff from among engineers who later got an MBA.

This is also how schools like Stanford and Berkeley set up a mini-business school (or “engineering management” program) within their engineering school. And it’s also why MIT recently set up its Engineering Leadership Program for undergraduates.

On the other hand, a number of schools run business plan classes and programs by assuming individual specialization and deliberately mixing the various specialists. The NSF-funded Georgia Tech Tiger program is perhaps the best known such program among those of us who teach tech entrepreneurship. To some degree, this reflects the supply limitations — you can’t get enough cross-functional people so you merge silo’d programs (with silo’d students) onto a temporary cross-functional team.

Obviously any good tech idea needs to be brought to market through a combination of technology, marketing, finance and manufacturing (or other operations) skills. How do you build such a team in a real startup? And who should be in charge?

I’m an engineer who went into B2B sales and marketing, so it’s easy to guess where my sympathies lie. And at a recent MIT club event on the “Gordon-MIT Engineering Leadership Program,” I heard veteran tech CEOs talk about how it takes a technical person to lead a tech startup.

Still, there are many counter-examples. For every Larry Page, there’s at least one (maybe more) Jerry Yangs.

Steve Jobs offers another model. Sure he was a great marketer — one of the best of the 20th century — and a great CEO. However, if you look at the recent Jobs biography, it was clear that his mechanic father and his childhood electronics experiments gave him an intuition about engineering design that many practicing engineers lack. (Alas, as the original Mac 128 death march demonstrated, he also had completely unrealistic expectations about how long things should take.)

So how do you form a cross-functional team to make the next great tech startup? And how do you allocate decision rights and authority among them? How does this change over the life of the firm, the industry and the technology? And what do you do with your hybrid business-engineers (or business-scientists)? They’re not going to be CTOs or CSOs (are they?), but do you put them as VP of R&D, or division managers, or CEO?

These are all interesting questions. Perhaps someone will research these answers.