It’s been a busy one. 2000+ applications, 300+ interviews, 53 Founders, 3% acceptance rate, 400+ ideas explored, 19 ventures built, all in really tough areas from digitally accelerated therapeutics to brain machine interfaces, 63% chance of a founder starting a venture, 20 top tier industry partnerships, 13 patents filed and 50+ leading academics working with us to build ventures from across universities from Imperial to Stanford. Here’s how we got there.
The journey so far
This time last year we had just quit our jobs and were camping in Imperial’s Incubator with a big dream to create the perfect environment to start a science company, and very little idea whether it was really possible.
As DSV has evolved over the last year this mission has become clearer. Namely that the biggest challenges facing humanity at the moment undeniably require the convergence of the sciences in a rapid and directed manner. One that more closely approximates the profile of an engineering or tech company.
Currently this is rare, and progress is slow, not just because science is hard, but more due to structural challenges and assumptions inherent across the value chain, more on that here. In order to create the future we want to see, we need a vastly more effective approach to the initial stages of science-driven company creation, and this continues to be our sole focus going into 2018.
Over the last year there’ve been more ups and downs than we ever anticipated but it’s now clearer than ever that we are on a very exciting journey. Building anything is hard, and we believe it’s best for the ecosystem as a whole to be open and candid (huge respect to SeedCamp for their efforts in this area). So here’s our year in review. No hype, just a passionate little company hopefully on its way to make a big difference in the world.
Luckily we weren’t new to science or venture, five of the companies that we worked with at Imperial have since exited to big names like Google, Facebook and Spotify. And whilst many of the early exits were pure digital ‘tech’ ventures we also had the pleasure of working with a wide range of ‘proper’ deep-tech companies from blood based diagnostics through to materials for non-volatile memory.
Many of these companies have since raised subsequent rounds and launched partnerships with major corporates to scale their offering. In most cases these companies started with a cutting edge technology but very rarely identified a real market need (i.e. tech-push). Therefore we were at least confident that our familiar approach to customer development could reliably yield an effective route to market even in some of the toughest fields.
The challenge was: we weren’t looking to launch an accelerator, or to find market fit for existing technologies. We specifically wanted to take an alternative approach because we were all too aware of the scaling challenges for a company that starts from a tech first position.
Rather, we believed that over the long term the most impactful companies will be created from a purpose driven perspective, with multiple technologies brought together to address a larger market led proposition. This might not sound strange but stands in stark contrast to the established approach in science, ‘technology push’, which principally focuses on the novelty of a specific technology (reviewed in a previous post). To do this we needed to find the right people, create the right environment and build from scratch. And this was way out of our comfort zone.
Fortunately the ‘venture builder’ model, i.e. a company that builds other companies, turned out not to be as novel as we first thought. There are actually hundreds of venture builders across the world, including of course some fantastic ones in London with EF and Founders Factory. Most venture builders are focused on Software as a Service, most stay small (2–5 companies per year), most don’t last very long, but those that do produce companies that do better than the usual odds.
A few however, are focused on science, or rather specifically biotech, again at a relatively small scale but creating some truly category defining companies. This includes companies like FlagShip Pioneering, Third Rock Ventures and many of the well-know funds such as Medicxi and Touchstone. Our hypothesis was that there’s a way of taking the approach used to generate biotech companies in these outstanding funds and make it scalable and applicable across the sciences.
We started with several leap of faith assumptions. We assumed that getting 20–30 entrepreneurial, creative, technically outstanding scientists and engineers in a room together was both possible and would constitute the set of core components required to build scalable, defensible, science-based companies. Moreover that, we assumed that through a network of facilities, we would be able to find these teams the equipment they needed to build early stage technical proof of concepts within reasonable capital constraints. And finally that, companies that crossed sector and disciplinary boundaries would be understood by investors.
Boom, there’s a reason people haven’t scaled science venture creation
Over the summer we interviewed hundreds of PhDs, PostDocs and industry scientists. The first big surprise was that the type of creativity, drive, growth mindset and greater awareness required to build a company is even rarer than we had anticipated. This kind of ability is obviously rare in the wider population, but we had assumed that in this class of Nature published, ox/cam/imperial/stanford graduated, big name start-up / big bank execs, more than 3% would fit the bill, we were wrong and would need to substantially widen the search. On the flip side, this degree of fragmentation cemented in our minds the need for an environment to bring these people together. With a wider search we found 30 incredible people and on December 10th 2016 kicked things off with a big party to bring everyone together.
Within the first few weeks we learnt that a principally customer development driven approach, which works so well for both tech-push and purely digital venture acceleration, really wouldn’t be sufficient to drive the kind of class-leading impact that we were aiming for.
It’s relatively easy to find an opportunity to increase efficiency of some process by 10%, find an off-the-shelf technology to fit and sign-up the first early adopters. Maybe, in time, that can provide a meandering route to more fundamental change, but so often that journey seems to result in ultimate indifference from the market, little differentiation vs. the competition or early exits as noone is really that commited to the mission.
Our aim is to find beachheads that enable a more direct route towards fundamental change from day one. Stitching together these kind of opportunities turns out to be much harder than anticipated, primarily because most industry execs think at the level of incremental change in a very specific domain (or their immediate problems). Meanwhile most investors think only in terms of current trends, rather than applications, or when pushed borderline ridiculous sci-fi futurism. There are people who can take a higher level view of an industry and pinpoint scalable opportunities but, again they are far rarer than we had imagined.
It was evident that without considerable insight into a given opportunity from all perspectives across the value chain, and particularly with regard to a deep understanding of the science, the tendency would be to drift towards one of two outcomes. Either very high-level propositions that would sound very impressive at first, but rarely survive through to scale in UK’s risk averse funding environment. Or super specific, more tech-push like solutions that at best would see early exits, but not drive the kind of change that we were looking to create. Our focus going forward would be to shift away from the typical accelerator production line and more towards this kind of deeply immersive environment in areas of high venture and impact potential.
Despite our initial design flaws, we arrived at results of which we are exceptionally proud. The founders took the programme’s imperfections in their stride, forming unconventional, multidisciplinary teams around smart approaches to fundamental challenges such as antibiotic resistance, drug discovery and creating completely new materials that address the toxicity inherent in all industrially applied adhesives.
In total, 400+ concepts were reduced to 18 proto-companies of which we invested into 13, plus one external company bringing the total portfolio to 14 ventures. The companies that didn’t quite fit our investment thesis have in the main (80%) have also gone on to raise external funding and we continue to work with them. At the current point in time around 80% of teams are ahead of schedule on proof of concept development whilst 20% have hit bumps and are working around them. Even more encouragingly, 70% have signed serious agreements (not LOIs!) with customers for trials and future purchases.
This was an encouraging start but we are well aware that many will no doubt fail over the coming years. From the rare data that is released, the failure / zombie rate at pre-seed with a high quality investor seems to sit at around 70%. Whilst we of course aim to change these odds, there is no doubt that 2018 will see many a trying time, especially given the scale of the challenges being undertaken and how this sits at odds with much of the UK’s funding landscape. Creating a route to scale past these hurdles will form much of the focus of this year.
The DSV portfolio as of December 2017
- Antiverse: one day de novo antibody design to accelerate the pathway towards a wide range of therapeutics
- Synbiosys (stealth): the world’s lightest, strongest and transparent shock-proof material
- Continuum: textiles seamlessly integrating high precision sensors
- Rapifage (stealth): distinguishing between bacterial and viral infections in less than 60 seconds in a way that actually integrates with the primary healthcare system
- MaterializeX: designing and optimising non-toxic adhesives
- uFraction8: high efficiency liquid particle separation to enable the bioeconomy
- P.E.S (stealth): Mapping the microbiome activity in agriculture to create a far more intelligent approach to increasing yield
- Scalpel (stealth): Aiming to drive the number of surgical errors globally to zero
- Infigear (stealth): Lighter, more efficient, longer lasting gearing
- Beta Bugs: Directed evolution of insects to create better food sources for aquaculture and livestock
- Immune Compass (stealth): Significantly reducing the IVF failure rate
- MicroInventa: Enabling the culture of unculturable bacteria to create new classes of therapeutics for chronic disease
- Neuroloom (stealth): High resolution brain machine interfaces, starting with restoring sight with retinal implants
- The Engineering Company (stealth): Turning years of heavy engineering into minutes of simulation for everything from turbines to rockets.*
*Technically not built internally but we’ve worked with Pari for years and were honoured to join the latest funding round.
Wrapping up year 2017 and on to year two
Finally, we want to say a huge thank you to everyone who’s helped to make this possible from our investors and sponsors (Imperial Innovations, CASTS, Wellcome Trust), to the new additions to the core team who could probably have a much easier life elsewhere (Santa, Hushpreet, Francis, Thane), and the many advisors across the early stage venture community and academic landscape who have generously given their time over the last year. Most importantly the 53 founders who took a punt that, despite ample other options, DSV was the right place to make a dent in the world. And if you’re reading this thinking you’d like to build something seriously impactful in the sciences this year then find how to join us here.