Unlocking The Promise Of Biotech Startups: Learnings From Mars, Goliath, and Evolution

Posted in Biotech investment themes, New business models, Pharma industry

As an early stage biotech investor, the future of the industry and its structure are clearly important things to think about – especially since most of the drug programs our new startups are working on won’t get launched, if they are the rare success, until well into the 2020’s.  A few recent experiences and observations prompted some deeper thinking, in particular around the Startup vs. Big Firm dichotomy and its implications.  I’ve boiled these down into three themes.

First, capital constraint “being the mother of frugal innovation”.  A few weeks ago there was a very interesting New York Times piece titled “From India, Proof That a Trip to Mars Doesn’t Have to Break the Bank, describing how India recently launched its Mangalyaan Mars Orbitor Spacecraft for only $75M, whereas NASA’s Mars mission, called Maven, cost $671M.  A cost differential of nearly 8x; this isn’t unexpected from the country that produced the $2.5K car and the $49 tablet, as the article highlighted.  The Mangalyaan was leaner in every way: carries only five instruments (vs eight), weighs half as much, was done in 18 months (vs the typical 72 months for a space vehicle) in order to catch the right alignment of sun, Mars, and Earth. They apparently had Gannt charts by the hour, not by the day or by the week.  Furthermore, they didn’t have time for multiple iterations and practice runs – make it good enough to get the job done the first time, work fast and furious 24-7, leap frog with learnings from others, and leverage the modularity that exists elsewhere in the ‘ecosystem’.  Despite the cost differences, both missions should be able to test key hypotheses around organic materials and possible life-supporting environments on Mars. Maven presumably asks more questions (its got more instruments!) but the fundamental questions are similar. As GE Healthcare’s Terri Bresenham says, quoted in the piece, “If necessity is the mother of invention, constraint is the mother of frugal innovation.”  While positive Indian efficiency was the gist of the article, an additional perspective might be on NASA’s Maven program – big bureaucracy, cushy government contracts (e.g., $3000 hammer anyone?), pork politics, costly regulations, and an overwhelming fear of failure presumably add a lot to Maven’s costs.

There are obviously huge parallels to the world of startup biotech vs big firms with this story.  Big budgets embody most of what Pharma does – every program tends to be built like a gold-plated Cadillac, saving a day on the launch is viewed as more important than de-risking each incremental dollar (here), and no one wants to be the project leader who failed to check all the boxes.  But capital constraints found in leaner companies mean you can’t do everything – you can’t have eight heavy instruments or fifteen different animal models.  In the drug business, as with Mars missions, appropriate capital constraints are clearly a positive force in most cases for doing things differently in companies of all sizes.

Second, common perceptions of advantages and disadvantages are often wrong. This came out of an hour listening to Malcolm Gladwell’s “David And Goliath” book on tape (yes, instead of reading it, I listened to it on the weekend’s drive back from skiing). The book goes through some well-trodden ground about underdogs and the advantage of being disadvantaged, with Gladwell’s superb storytelling narrative. Although I have my quibbles with some of the content, the general principles were framed in a thought provoking way: how the advantages of the strong are the root of their weaknesses, and vice versa.  New ways of looking at situations offer a contrasting view of strengths and weaknesses.  Related to the title, Gladwell reframes David as a nimble, expert projectile-thrower and Goliath being an over-armored infantry soldier, nearly blind and suffering from acromegaly.  This flips the conventional wisdom about who one would have expected to win on that battlefield in the desert of Palestine.

Applied to industry, this is classic Innovator’s DilemmaAttacker’s AdvantageCreative Destruction type stuff.  Pharma’s perceived advantages – more resources, more scale, more reach, more integration – might be anything but advantages.

Third, intelligent design only works in a world without evolution. I got thinking about this topic while reading some rubbish on the ID (intelligent design) movement. For me, the human body as a consequence of evolution is Exhibit A in the case against any intelligence in the process. The human body and its failings are clearly the product of a path-dependent evolution; for instance, we are burdened with hip, knee, and ankle joints that largely evolved before we became bipedal species (hence the lower back pain and joint surgeries), and we have at least ten vestigial organs with no apparent use today that we as organisms are forced to grow and maintain over time – like our tonsils, appendix, coccyx, and wisdom teeth, to name a few. Even goose bumps are vestigial – think about all the wasted calcium handling and contractile protein inside your skin’s arectores pilorum muscles to create those bumps. If we were intelligently designed, we’d have no useless body parts and evolutionary leftovers. We’d also have more bipedal-equipped joints. I’d propose that we’d have some built-in and improved redundancies to prevent single codon changes from causing catastrophic developmental impacts or cancerous predispositions.  Clearly, we aren’t made in the image of anything perfect – but instead constructed over the course of millions of years of evolution, from bugs to multi-celled organisms, from the Cambrian explosion to furry small mammals, from apes to multiple species of hominoids, and finally to today.  We carry tons of integrative viruses, pseudogenes, junk DNA (though not all is junk), pieces of Neanderthals, families of overly homologous proteins, and layers upon layers of regulatory pathways – all of which presumably derive from the mutation-, gene-copy-number-, and infection-driven molecular evolution of the human genome.  We are products of our evolutionary path, not intelligently designed organisms.

Just like humans, Pharma companies today are the products of path-dependent evolution as well.  The industry has evolved over the past fifty years from a near cottage industry of drug hunter champions into a model of largely centralized, bureaucratic R&D of both significant and suffocating scale; I’ve railed on about this subject in the past (herehere) so will try to avoid (unsuccessfully) repeating myself.  If you were designing a Pharma from scratch, would it look like any of the big firms today?  Most certainly not.

Reflections on Pharma

Stepping back, all three of these themes – constraint being a productive force, advantages being disadvantages and vice versa, and path-dependent organizational evolution – resonate deeply with my perspective on the Pharma industry today. To illustrate further, and explore these themes perhaps beyond the asymptote of their usefulness, here’s a list of a few of the perceived advantages of Pharma R&D today:

  • Almost unlimited access to all the latest technologies across drug discovery, ADME, toxicology, and clinical development, including all the latest capital equipment, compound libraries, antibody approaches, etc
  • International reach to support global clinical and regulatory processes to fully enable drug development programs
  • Deep and insightful commercial input into the markets, the pulse of the practicing physician, and the payors on what’s the right product profile
  • Gigantic cash flow streams that provide 15-20% of the topline to support a largely “block grant” model of R&D (fixing R&D spend to the percentage of sales)
  • Decades of institutional memory providing the scar tissue around what works and what doesn’t (e.g., insight into project attrition at massive scale)

This is a solid list of advantages, and they all have real merit.

But like the biblical Goliath, whose size and strength appeared to the Israelites as great advantages, they are also the roots of Pharma’s disadvantages.  All of these derive their value as inward and relatively insular forces.  Institutional memory in particular can serve to either unlock better paths to innovation or to stifle those that want to explore new ways of doing things.  Lipinski’s Rules, hERG liabilities, and other candidate guidelines derived from legacy “survivor bias”-style analyses are case examples of this tension – unfortunately the stifling aspects rather than the unlocking ones often triumph in big firms.

Further, these impressive corporate R&D “advantages” are of course the product of Big Pharma’s path-dependency: single blockbuster successes discovered in the ‘60s-70s led to early mergers in the ‘80-90s, and bigger mega-mergers in the late 90s-00s, to form the organizations of today.  Bigger and bigger R&D budgets buying up more and more “things” in the quest for improved productivity. In a sense, the growth drivers underlying these mergers acted like the excessive hGH coming from Goliath’s pituitary – the scale and constant growth pressure was a product of a disease, not a design.

“Organizational acromegaly” created a body full of the legacy excess infrastructure, technology, and staffing all in the hope of achieving more launches.  But the organizations couldn’t cope, and couldn’t deliver. Plagued further with institutional baggage of this corporate evolution, innovation suffered.  Culture went sour.  What may have been an advantageous feature at one time in Pharma was now a disadvantageous bug causing havoc and stifling new discoveries.

Today, trapped by the scale of their in-house technology platforms and global R&D staffs, Pharma organizations are forced by inertia to feed them with more projects and capital, or face frustratingly idle capacity.  These legacy costs can bleed the budget of an organization for a long time before they are forced to make tough decisions. Global footprints with multiple international R&D mega-sites and high fixed infrastructure costs are a key part of this budget bloodletting.  The closure of many big R&D sites in the past couple years shows how Pharma leadership has come to the conclusion that the only way to cut costs is to cut infrastructure – you can’t meaningfully dent the budget in any other way.  Cost constraint may be the mother of frugal innovation, but huge legacy fixed costs make frugality hard to imagine.  Just turning the lights on is a major expense.  Some big firms are trying to externalize access to these platforms, to share their capacity with the ecosystem, but only time will tell if they can get out of their own way in order to make it happen.

Things are clearly changing over the past few years, but the biggest obstacle to change is the costly inertia of legacy institutions – the governance, processes, structures, and cultures that, like evolution’s integrated viruses and vestigial genes, are very hard to remove from the organizational genome (see here for thoughts on how to change these).

The Power and Potential of Startups

Startups benefit from the convergence of all three of these themes.

At their inception, startups offer a clean sheet of paper for the founding team – freed from legacy encumbrances and infrastructure, startups can actually engage in real ‘intelligent design’ of an R&D organization, embracing more revolution than evolution.  They don’t need to be – and shouldn’t be – created “in the image” of Pharma.

They can reframe the model – operating outside the legacy establishment in unique ways with other players across the globe, creating distributed, highly networked R&D systems.  Not having access to any existing capabilities or technology footprints may be a perceived disadvantage, and if poorly navigated it certainly is, but it can also open up new collaborative models and partnerships.  The latter are real advantages that feed off the strength of the growing networks in the life sciences.  Our portfolio of drug discovery seed-stage startups just reviewed the scores of CRO partners they are working with, and possible risk-sharing models, that enable innovation in ways inconceivable only a decade ago.

Like India’s space agency, they can tackle big projects while leap-frogging forward, taking advantage of capabilities and modularity developed elsewhere in the industry and now available for all across the ecosystem.  By its very nature, a startup’s de novo process liberates teams to consider leaner, better ways of doing things; by its design, this process allows positive cost constraints and productive capital efficiency to permeate the very essence the new enterprise.

A biotech startup created in this manner doesn’t look like a lowly shepherd boy walking up to fight a huge warrior with only god on his side – it looks like a lethal, well-trained projectile thrower ready to bring down a slow, over-burdoned and diseased opponent.

I often describe the creation of a new startup as one of coding the new DNA of an organization: starting from scratch, we can avoid known mutations in key governance genes, make sure several copies of the key scientific genes are replicated, express significant talent-recruiting chemotactic agents early in the life of the company, etc… One can take the analogy too far, and I often do.  But fundamentally, ‘Intelligent Design’ of the DNA of a startup is doable because we can throw away the legacy corporate genome.  We can legitimately ask, if we were to design the drug discovery entity of the future, how would you do it?

Figure out what you want the organism to do, and design the “minimally complex genome”, in synthetic biology speak, to achieve that goal.  Are you building a single-asset virtual company – there are a set of genes for that (see key success factor genes here).  Are you creating a new drug discovery platform or biology-led product engine – there are another set of genes for that (here). Early in the life of a startup organism, optionality can be preserved by waiting to express genes later in development – preserving the ability to build a platform or focus on a single project after a business model signal has declared itself.

Freed from the baggage of organizational evolution, startups can code their own DNA, reframing disadvantages into advantages, creating unparalleled networked models, and embracing frugal but impactful innovation.

The era of ultra nimble, newly coded biotech startups is just beginning – welcome to the revolution.

This entry was posted in Biotech investment themes, New business models, Pharma industry and tagged , , , , , . Bookmark the permalink.
  • ayatollahOfTheOutcomes

    Interesting blog! You fail to mention the role of Legal in big pharma in quashing every bit of innovation. Patents don’t get traded as often as they should (think of all the early stage projects – the lawyers won’t let the company reveal anything about the state of the program to the competition – if they do decide to, BD makes it a very expensive exercise to do any discussions around selling assets such as patents, data etc.) New technologies get quashed for fear of opening the co. to lawsuits. Risk-Benefit Analysis doesn’t exist in their analyses. Also failures and data on failures are rarely if ever, shared All of this has led to the outcome morass that pharma is experiencing today.

  • CMCGuy

    You pose the question “If you were designing a Pharma from scratch, would it look like any of the big firms today?” and while indeed would not advocate adoption of culture of any of the current organizations many of their progenitors from a few decades ago prior to merger mania and profit centric-ism might be more suitable models then reliance on start-ups, which do have a great many disadvantages in practice. There used to be multiple medium to large pharma companies that were both focused on areas and programs while having enough critical mass and right expertise to discovery and develop drugs. There was balance in doing innovation and having more complete capabilities without the heavy infrastructure burdens that have grown. Perhaps we can not go back and maybe a few of the start-ups can evolve to be more than just nimble which can likely only go so far.

  • Steve Betz

    I like your use of the David-Goliath inversion of perception and expectations, but I’m surprised by your Mangalyaan vs Maven comparison — especially by someone that’s written so convincingly about the power of the Team.

    As a venture, Team Mangalyaan has little to no experience and no success in this area, while Team NASA has literally decades of success (Mariners, Voyagers, Vikings, Cassini, Mars Global Surveyor, Pioneers, Spirit, Opportunity, Curiosity) with very few ventures failing (see Jade Rabbit for another untested Team’s recent failure). MAVEN has 4x the capacity and is expected to last 2x as long for 8x the cost. That’s not saying that you couldn’t save some $ at NASA but the numbers don’t seem particularly out of hand.

    As an investor, would you put your financing behind Team Mangalyaan (high risk, potentially higher ROI) or Team NASA (low risk, steady reward)?

  • LifeSciVC

    Fair enough, and good counterpoint on the details. My point was more around the general rather than specific observation that big budgets often lead to a “build the most expensive machine in the hospital” (Monty Python reference) approach rather than focusing on addressing the key risk-reducing question in a focused and efficient way.
    Sent from my iPad

  • Steve Betz

    I completely agree that many organizations seem too willing to go “all-in” only and don’t try to support/disprove projects surgically (and early). I’ve worked in ones where once it was “go”, it was almost impossible to pull the plug because losing-face was unacceptable. Nimble, focused orgs MUST avoid that trap.

  • http://thedivelab.com Sebastien Murat

    Bruce.

    Though I generally agree with your section on ID, those non-ID vestigial `errors´ are small enough not to really break one’s back. Also, ID does not necessitate that we be perfectly or even, indeed, well adapted, say, specialized to our environment. Going without the surplus, and lean, i.e., specialized might be intelligent in in a very stable & predicatble environment, but hardly in an increasingly dynamic & unpredictable one. Actually, it could be argued that the opposite is true, i.e., having a sufficient amount of `junk´ seems intelligent in the face of an uncertain future, especially one undergoing accelerated change, since even evolution can’t see around the corner. Think of it has a junk-yard offering unmatched capability for bricolage, on the run or, for an analogy you might be able to easily relate to, preparing for a camping trip. I know that more than half the stuff I take camping I don’t need (inexperienced camper? smart camper?). Now, I don’t know for sure which things I surely will need, basics aside, so I take as much as I can (the `you never know it could rain or we might be attacked by some wild animal´ thingy). Of course, some things on any trip prove useless, maybe even render the trip downright uncomfortable, maybe even giving me a sore back, and all the more on the way back because I wasn’t in a position to throw it all away, for I might need it on the next trip. In such case, selecting for things with redundancies (evolvability) seems awfully like ID. Indeed, there exists evidence of life selecting for evolvability.

    The link below of a picture story adventure I experienced shows how those vestigial remnants needn’t be all bad, could actually be quite good, maybe even very, very good in terms of offering an in-the-distant-future optionality & avoid going extinct, i.e., not just at the organism level, but at all levels, including, as is the case in point, an übermodel of a start-up biotech built for unmatched, say, omnipotent optionality.

    Once again, thanks for all your stimulating pieces, like gems in a junk-yard!

    http://www.slideshare.net/SebastienMurat/hibernaut-human-20doco-pitchdeckthe-dive-lab

    PS: it’s best read as a bed-time story :-)