DeepTech: Many Roads May Lead There, But There’s Only One Rome
Today: A guest post by Bill Janeway, veteran venture capitalist, author and faculty member at Cambridge University.
Back to deeptech! Last year I wrote a piece to complain that we didn’t have enough controversies in European tech. That’s a problem because I think intellectual battles greatly accelerate the advancement of ideas.
And then in December we did get a controversy, about the opportunity that deeptech represents for European entrepreneurs and investors. It started with a column by me in Sifted (inspired by Jerry Neumann’s work on Productive Uncertainty being the prime lever of success for tech startups). Here’s the gist of it:
Too often, European founders and investors express resignation in the face of the disappointing performance of local tech companies when compared to their counterparts in the US or Asia. And they come to the conclusion that because European startups have a harder time scaling up (because of the continent’s characteristic market fragmentation), local founders should instead focus on ‘deeptech’ challenges.
A few days later, Zoë Chambers of Octopus Ventures, who focuses on deeptech startups, struck back with a column in which she makes the case for deploying more capital in deeptech—to be expected from the part of a deeptech investor, but I really do appreciate it when VCs articulate and discuss their thesis in public (we don’t have enough of those in Europe). Have a look at Zoë’s column:
Innovation isn’t linear, and what people might find useful, desirable and eventually essential isn’t known until the possibilities are explored. When we invested in Evi (then True Knowledge) and its ‘answer engine’ back in 2007, we were investing in uncertainty. Not by turning away from deeptech, but by appreciating that the uncertainty created by this new technology could unlock a big opportunity and, in this case, Amazon’s Alexa was born.
I then forwarded both columns to Bill Janeway, a veteran of venture capital, economist and author of Doing Capitalism in the Innovation Economy—a book that taught me almost all I know about technology and venture capital!
Bill was kind enough not only to share detailed ideas and references, but also to agree that his thoughts could be published as a guest post in this newsletter.
There are numerous, interwoven threads to be picked out here.
I continue to believe that at any point in time there is more technology than anyone knows what to do with and that merely adding to the stock of technology – with no identifiable market opportunity – does not create value.
Thinking this way originated in my experience of a unique conjuncture: the transformation of computing in the 1990s, from the vertically integrated, closed, and proprietary world of IBM and the 7 Dwarves and DEC and Data General and HP to the horizontal world characterized by open interfaces, published protocols, and distributed computing resources.
That time – the 1990s – was characterized by a reservoir of orphaned technology, developed by obsolescent incumbents as they attempted to make the transition: VERITAS and BEA were built by selectively picking over the orphans and acquiring the IP through startups led by visionary technologists who also knew both the relevant markets and how to build the channels to those markets.
However, by 1997-8, when distributed computing went exponential with the incipient commercialization of the internet, the needed new technology had to be built: it was not available to be bought (although I managed to join with BEA’s founders to get the crucial acquisition of WebLogic through a skeptical board because Warburg Pincus [Bill’s firm] still owned about half the company, see Doing Capitalism, pp. 137-9); so the excess of available technology now is largely the work of entrepreneurs and VCs competing to add functionality and therefore value to a visibly evolving environment.
But the “deep technology” that has continued to enrich and extend the digital environment has emerged both from established incumbents (Amazon’s suite of cloud resources, for example) and from startups (from Salesforce to Databricks); the “Division of Innovative Labor” identified by my friends at Duke (A. Arora, Cohen, W. M., and Walsh, J. P., The Acquisition and Commercialization of Invention in American Manufacturing: Incidence and Impact, Research Policy (July 2016), 45(6), pp. 1113-1128) suggests that startups take on more of the “exploration” risk while incumbents, take on more of the “exploitation” risk, often by acquisition (this may relate to Neumann’s invocation of “uncertainty” which to me seems rather untethered).
One challenge for startups doing deep digital tech for the enterprise market has been the shift in the revenue model from proprietary license to SaaS (Doing Capitalism: Enterprise Software Transformed, pp, 142ff.—also available here as an article), which requires vastly greater external capital to fund growth since the customers no longer are paying upfront: this is not a problem as long as the Unicorn Bubble lasts, but it does increase the weight to be put on financing risk relative to technology risk and market risk (see Nanda, Ramana and Matthew Rhodes-Kropf, Financing Risk and Innovation, Management Science 63, no. 4 (April 2017): 901–918).
I see nothing irrational about entrepreneurs setting out to deliver digital deeptech innovation with the aspiration either to reach the promised land of positive cash flow from operations (rarely) or to be acquired along the way (less rarely), or to fail (most likely); whatever the outcome, they are emulating, as you say, biotech entrepreneurs and VCs.
The burden of financial risk interacting with market risk is greatest in the domain of digital services for consumers, where only persistence of bubble financing is likely to fund startups to sufficient scale to attract advertising revenues, especially if the incumbents are effectively barred from acquiring potential competitors.
Of course, progress at the frontier is made by trial-and-error experimentation, and one recent development is that in the digital domain, the cost of experimentation has fallen drastically (see M. R. Kerr, Nanda, R., Rhodes-Kropf, M., Entrepreneurship as Experimentation, Journal of Economic Perspectives, 28:3, Summer 2014, pp. 25-48); in turn, this means many more start-ups developing “too much” technology with varying clarity of insight with respect to possible target markets: this is “Schumpeterian Waste” in action!
What is implicit in your exchange with Zoë Chambers [here’s Zoë; here’s Nicolas] and should be made explicit is the different weights to be put on technology risk versus market risk: successfully resolving market risk does appear to generate more reward than resolving technical risk: many roads may lead there, but there’s only one Rome!
I apply this to the BioTech paradox: why have VCs and public markets investors alike poured countless billions of dollars into ventures with enormous technical risk? Answer: because if the technical risk is overcome, there is no market risk (third party funding of quasi-vertical demand curves: see Doing Capitalism, pp. 96-101).
This is one reason why I put so much emphasis on the state’s role as the early, supportive customer for frontier technology, mobilizing the procurement function to eliminate market risk before the new technology is commercially ready while pulling the suppliers down the learning curve to low-cost, reliable production: the model for the semiconductor industry in the 1960s/1970s is available for the energy storage industry in the 2020s!
Enough for now….Hope this helps.
Here are various articles referred in the discussion or that directly relate to it:
Unicorns: Why This Bubble Is Different (William H. Janeway, Forbes, May 2015)
From Atoms to Bits to Atoms: Friction on the Pathway to the Digital Future (William H. Janeway, Forbes, July 2015)
Enterprise Software: Death and Transfiguration (William H. Janeway, WTF?, October 2016)
The Hard Truth About Deep Tech (me, European Straits, June 2019)
What Is Venture Capital? (me, European Straits, March 2020)
When Tailwinds Vanish (John Luttig, Luttig’s Learnings, April 2020)
Does Every Country Need Their Own DARPA? (me, European Straits, October 2020)
Productive Uncertainty (Jerry Neumann, Reaction Wheel, November 2020)
Is investing in 'deeptech' a good move for Europe's VCs? (me, Sifted, December 2020)
Europe needs to back more deeptech startups (Zoë Chambers, Sifted, December 2020)
📘 In addition, here’s Bill’s book on Amazon (2nd edition): Doing Capitalism in the Innovation Economy: Reconfiguring the Three-Player Game between Markets, Speculators and the State.
If you enjoyed this edition of European Straits, you should subscribe so as not to miss the next ones (special offer until Sunday, January 10) 👇
From Munich, Germany 🇩🇪