Interesting piece! The types of companies you talk about fostering make me think of the the German Mittelstand companies, which are often highly technical, innovative and focused on dominating niche markets. They also create mutual interdependence between the company and its employees. I think the American way of creating would be employee options, which it seems to me are underused outside of the startup/tech world.
Really like the direction of this piece. I think you could make it even better by condensing it into shorter form and then spending more time trying to make concrete what your recommended approach would be. I am picking up three threads… A. having companies be more profit focused or at least have a mix of research and profitable projects. B. Have universities try to be more revenue generating via their research. C. Have corporate work more closely with universities, students, and projects (although the blocker here seems to be on the university side and incentives).
Perhaps another argument is that VCs need to be less diversified, perhaps even to the extent of the VC being CEOs or operators of the companies or at least being more closely involved . The problem for VCs with deep tech is that the returns are just lower and maybe the only way to combat this is to reduce the variance in returns by being more focused (although from a population viewpoint, there will obviously still be variance, and I’m not sure if this ultimately improves the risk reward ratio). But maybe the variance of returns from deep Tech companies – at some level – is indeed lower and that is where the opportunity lies in getting that better ratio by reducing diversification.
Re: “Working more closely to help tailor projects the VC is not currently funding to be more VC-investable in the future” - check out Conception X (UK-based) https://www.conceptionx.org/
I like this concept. It strikes me that ESOPs and worker ownership would probably fit well as an exit option for these William Walker-style funds, given that the idea is that they are generating cashflow almost from the beginning.
Interesting writing and ideas! Nonetheless, as a deep tech VC allow me to tell you that, per your last lines on recommending us how to allocate "a massive amount of time", we tend to be good at telling a founder if their idea is good or bad, and investing in it in consequence. However, we are not usually good at producing innovative ideas of our own; much less to feeding them into academia to steer research topics. That may be a work better performed by incumbents or customers, who are more directly in contact with the several pains to solve in an industry. Another story is to be asked about concrete research subjects and what their expected impacts in industry could be.
Great piece! One big reason that I think any sort of deep tech VC won't be successful is that most of the economic profit derived from deep tech innovations is not actually captured by the inventors. Bell labs invented the transistor, and they captured an epsilon of the profit from the subsequent creation of the computer industry. As you note, any VC will have to acknowledge that investing in deep tech will be a worse proposition than normal other investments, and they will have to be OK with some other companies will take advantage of the inventions in which they invested to make many multiples more than they will.
I mean, I guess I have two main things to say on that.
The main one is that Bell did not really 'try and fail' to profit from the transistor. They had a decades-long, ongoing conversation with the antitrust folks in DC and knew that they were keeping a very close eye on Bell -- particularly on how horizontally integrated they felt Bell was becoming. So when something like the transistor sprung up in Labs, nobody knew better than Mervin Kelly that they were going to have to license that patent for free and maybe just get it out of Bell Labs as fast as possible.
The second is just that everyone in many kinds of tech industries throughout history both has a strong tendency to use trade secrets where they can and doesn't tend to capture much of the value they create. The existence of the possibility of IP theft and creating some non-internalized social good has always been the state of things. I'm not sure that constitutes a good, coherent reason for an area of investment to not exist in and of itself.
This is a really great analysis of market differences between software and hard tech! Agree that there isn't really any institution out there that's doing this yet, but I think VC's are definitely experimenting with new models for deep tech. Several that might be of interest and are similar to some of the tamer takeaways (and would be curious for reactions):
- Activate (supports scientists for two years to develop ideas, with funding and national lab support) https://www.activate.org/
- Nucleate (community-run, non-equity, startup accelerator for bio PhD students) https://nucleate.xyz/
Agree that none of these models have that exact combination of pre-problem discovery entry for graduate students and institutional funding/support for actually solving engineering problems first through consulting that you describe though.
Interesting piece! The types of companies you talk about fostering make me think of the the German Mittelstand companies, which are often highly technical, innovative and focused on dominating niche markets. They also create mutual interdependence between the company and its employees. I think the American way of creating would be employee options, which it seems to me are underused outside of the startup/tech world.
Really like the direction of this piece. I think you could make it even better by condensing it into shorter form and then spending more time trying to make concrete what your recommended approach would be. I am picking up three threads… A. having companies be more profit focused or at least have a mix of research and profitable projects. B. Have universities try to be more revenue generating via their research. C. Have corporate work more closely with universities, students, and projects (although the blocker here seems to be on the university side and incentives).
Perhaps another argument is that VCs need to be less diversified, perhaps even to the extent of the VC being CEOs or operators of the companies or at least being more closely involved . The problem for VCs with deep tech is that the returns are just lower and maybe the only way to combat this is to reduce the variance in returns by being more focused (although from a population viewpoint, there will obviously still be variance, and I’m not sure if this ultimately improves the risk reward ratio). But maybe the variance of returns from deep Tech companies – at some level – is indeed lower and that is where the opportunity lies in getting that better ratio by reducing diversification.
Re: “Working more closely to help tailor projects the VC is not currently funding to be more VC-investable in the future” - check out Conception X (UK-based) https://www.conceptionx.org/
I like this concept. It strikes me that ESOPs and worker ownership would probably fit well as an exit option for these William Walker-style funds, given that the idea is that they are generating cashflow almost from the beginning.
Interesting writing and ideas! Nonetheless, as a deep tech VC allow me to tell you that, per your last lines on recommending us how to allocate "a massive amount of time", we tend to be good at telling a founder if their idea is good or bad, and investing in it in consequence. However, we are not usually good at producing innovative ideas of our own; much less to feeding them into academia to steer research topics. That may be a work better performed by incumbents or customers, who are more directly in contact with the several pains to solve in an industry. Another story is to be asked about concrete research subjects and what their expected impacts in industry could be.
Great piece! One big reason that I think any sort of deep tech VC won't be successful is that most of the economic profit derived from deep tech innovations is not actually captured by the inventors. Bell labs invented the transistor, and they captured an epsilon of the profit from the subsequent creation of the computer industry. As you note, any VC will have to acknowledge that investing in deep tech will be a worse proposition than normal other investments, and they will have to be OK with some other companies will take advantage of the inventions in which they invested to make many multiples more than they will.
I mean, I guess I have two main things to say on that.
The main one is that Bell did not really 'try and fail' to profit from the transistor. They had a decades-long, ongoing conversation with the antitrust folks in DC and knew that they were keeping a very close eye on Bell -- particularly on how horizontally integrated they felt Bell was becoming. So when something like the transistor sprung up in Labs, nobody knew better than Mervin Kelly that they were going to have to license that patent for free and maybe just get it out of Bell Labs as fast as possible.
The second is just that everyone in many kinds of tech industries throughout history both has a strong tendency to use trade secrets where they can and doesn't tend to capture much of the value they create. The existence of the possibility of IP theft and creating some non-internalized social good has always been the state of things. I'm not sure that constitutes a good, coherent reason for an area of investment to not exist in and of itself.
This is a really great analysis of market differences between software and hard tech! Agree that there isn't really any institution out there that's doing this yet, but I think VC's are definitely experimenting with new models for deep tech. Several that might be of interest and are similar to some of the tamer takeaways (and would be curious for reactions):
- Deep Science Ventures fully funded PhD program (get your phd through doing problem discovery with a VC) https://deepscienceventures.com/venture-science-doctorate
- Fifty Years PhD to VC night-class basically https://deepscienceventures.com/venture-science-doctorate
- Activate (supports scientists for two years to develop ideas, with funding and national lab support) https://www.activate.org/
- Nucleate (community-run, non-equity, startup accelerator for bio PhD students) https://nucleate.xyz/
Agree that none of these models have that exact combination of pre-problem discovery entry for graduate students and institutional funding/support for actually solving engineering problems first through consulting that you describe though.