Creating 200 companies over the next 5 years
At the centre of our mission at Deep Science Ventures is the belief that if we are to tackle the existential challenges facing humanity, we have to see drastic change at the sector level across climate, agriculture, pharmaceuticals, and computation. To catalyse this change - and to do so at the rapid rate that is required - we have to fundamentally rethink how we approach venture creation and importantly, how we structure our broader innovation infrastructure.
In the past five years, we have aimed to do just that; we have created 35 companies in fields as diverse as direct air capture of CO2 to reforestation using mycorrhizal fungi, in-vitro antibody production and the creation of a universal compute fabric. Over the next 5 years, we aim to create 200 companies, releasing our toolset to the scientific community with a view to creating a truly collaborative venture creation ecosystem.
Behind our venture creation approach is a defining focus on ‘working backwards’ from outcomes, and meaningfully expanding on concepts such as bottleneck analysis, roadmapping, and tech trees. How it differs is that DSV starts with the desired outcome, which is identified across a wide community of around 500 advisors as well as the key input of our partners. Then we identify the technical or economic constraints that must be overcome to deliver the optimal approach - a process we call scoping.
This in turn is enabled by a unique ontology, which allows us to collate knowledge and encode first-principles thinking, making information easily understandable and shareable across all those we work with.
Starting with the Outcome
What our methodology means in practice is that we start with the desired outcome for an area, and work backwards to identify the potential approaches to achieve it. Our venture creation is based on those approaches. This refers to a vision of macro change, rather than a single idea or solution, based on which we recruit Founders with specific skill sets.
Our approach encompasses a variety of significant additional elements or Outcome “nodes” including constraints, solutions, hypothesised constraints, and hypothesised solutions, which we use to piece together a picture of what is feasible and desirable and the optimum ways to get there. We are technically agnostic to how that problem is solved, and that is crucial - this is not a controversial point in any other field of entrepreneurship.
While this might seem fairly intuitive, it runs counter to the conventional wisdom in the science venture world, with nearly all ventures starting with a new discovery or technology that must find a home in the world, or by asking what the expertise of a specific individual might achieve. Many science companies are spun out of university labs by tech translation offices, with the top 5 patent-producing universities in the US producing one-third of new biotech companies. However, this tech-push approach is far from a guarantee of success: 47% of those companies have already failed.
Our outcome-first methodology flips that approach on its head. This might be best illustrated by an example of how our ontology uses negative emissions technology. Our climate sector is focused on reversing global heating. We’ve identified a number of different constraints here, such as the need to remove gigatonnes of carbon dioxide from the atmosphere by 2050. Following that process, we created a number of negative emissions technologies companies such as Mission Zero Technologies, Parallel Carbon and Holy Grail, as solutions to hypothesised constraints such as the cost of electricity, supply of raw material and modularity.
Our work has also uncovered additional areas, such as the supply of critical materials, and the decarbonisation of industry (with their own internal set of constraints), that we’re now scoping and developing solutions for. Most importantly, our Outcomes Graph tracks this knowledge, with the ability to map shared constraints, even in areas where that otherwise wouldn’t have been clear. It should be noted that it is an evolving tool and we are looking to add more features and functionalities to it, which we believe will enable the collaborative paradigm of science that we are envisaging. You can read more on the technical working of our approach here or how we developed our approach to science venture here.
Our methodology allows our team to get just far enough, to understand the broad area of the most likely solution and optimum skill sets required to develop it. Once identified, we recruit a Founding Analyst to take it forward, eventually identifying the optimum approach to the solution, recruiting a founding team and spinning out as a venture. The quality of the companies that we create is therefore inherently tied to the depth of our knowledge of the areas that we build in, and our ability to identify technical and economic constraints within that field or market.
Core to our approach is our focus on developing the Outcomes Graph - effectively a new, evolving language that enables people to quickly get to first principles and communicate across domains in a rapid, programmatic way from the macro to the molecule. Through the Outcomes Graph, our outcomes-first methodology becomes reflected in the way in which our teams ideate, iterate and communicate. The tool also enables us to move fast, avoid biases and unambiguously explain why a given approach has the highest chance of succeeding. It is already leading to very promising results including a track record of 100% conversion from technical plan to lab results.
This stands in contrast to existing and prevailing systems that track knowledge across the sciences which, despite enabling easier knowledge mapping and search, have little or nothing to say about which and how this knowledge could be optimally applied to specific societal outcomes. That is something we aim to address.
Benefitting from the methodology
Ashraf is an ex-Intel Vice President, former Director at Bell Laboratories, and founder of Enpirion, a fabless semiconductor company, which was acquired by Altera Corp. With decades of experience in computing, he has joined us as a venture partner in our Computation sector, to identify constraints and accelerate our venture creation efforts within the sector.
Ashraf gravitated towards DSV specifically because of its methodology. DSV’s ontology builds upon his earlier experience conducting research at Bell Labs which used a similar approach that he describes as “another way of thinking about an outcome and finding different paths to get to that outcome”. Ashraf explains that this essential process involves bringing in people with a different perspective who can “look at the problem from other angles” and ask necessary and constructive questions.
In his eyes, this is part of what marks out DSV from other investment organisations. He says DSV’s methodology - especially the “deliberate scoping”, and the process of identifying promising areas at the outset - stands in contrast with that of other VCs, which he describes as often “superficial”.
“What's different with DSV is mainly in the preparation that happens upfront: The focus on understanding what the outcome needs to be, the possible ways to achieve that outcome and therefore, the specific skill sets and technical expertise of founders that would fit the bill and get you to that outcome.”
He adds that the Outcomes Graph is a “really good way to organise your thoughts and research areas, and especially obstacles and barriers so that you have a very complete picture”.
According to Ashraf, a core benefit of the outcomes-first methodology lies in focusing the attention of the founder and clarifying their thinking. “If you're in an area that maybe doesn't have an outcome at all, your problem is not that well defined.”
Kate, formerly part of the research team at Palo Alto Research Center (PARC), who is now at DSV building a company using abandoned pit mines to sequester carbon dioxide, has experienced similar benefits. She credits DSV’s “graph-based thinking” for helping her generate the central idea behind her venture.
She has also used DSV’s methodology to structure and stimulate her thinking, adding that it “revealed things that I definitely would not have come up with without it”. That in turn gave her both peace of mind and confidence in the choices she ended up making, reassuring her she’s not missing something major.
Our Outcomes Graph today sits at the very centre of our venture creation efforts, allowing us to create networks of knowledge across domains, as well as communicate them effectively.
Over the next few years we intend to build out our platforms to allow for enhanced collaboration, as well as allow for more powerful analytics, including augmenting the identification of potential opportunities for venture creation.
Ultimately, our ambition is to open up our platforms to more efficiently identify knowledge combinations with a higher probability of achieving given Outcomes across disparate domains, working towards a future in which expertise, knowledge and capital are automatically deployed and directed around a given Outcome.
We’re already using the ontology and Outcomes Graph in all of the roles that we’re currently recruiting for. If this is a topic that you are passionate about, please reach out to Eirini Malliaraki, who is leading on development at DSV.