Shaun Abrahamson is the managing partner and co-founder of Third Sphere, an early-stage fund investing in rapidly scalable technologies for climate change adaptation and resilience.
How did you get started in climate VC?
My background is in a bunch of different engineering disciplines—electrical, mechanical, and some software. I got my start during the dot com boom, which was a very useful experience to see how a bunch of really smart and thoughtful people are still just people in the sense that they can become completely irrational in the face of rapidly changing technology. About a decade-and-a-half ago I was still investing in early stage software, but I was also working with companies like Starbucks on sustainability-related projects. At some point, it became clear that although there was a lot of good will in the corporate environment, the startup model was probably the way to go if you wanted to do something about climate change fast and at scale. So around 2013, I began investing in climate with my co-founder Stonly Blue.
It was clear that very few of the things we needed could just be software so started looking for co-investors for hardware startups because we didn’t have enough money to do it ourselves. We made a bunch of investments in our first two years and those companies couldn’t raise at all. We really sucked at helping them because we assumed a lot of our software fundraising playbook was portable. We needed to spend some time trying to figure out how to package hardware investment opportunities so we could convert more investors. I still don’t think we’ve completely figured it out, but the climate problem for us quickly became a “how do you fund hardware” problem.
Why is hardware such a hard sell for VCs?
When we started Third Sphere, the question we were wrestling with was what people meant when they said that something was “capital intensive,” which was usually the reason they didn’t want to invest in a hardware startup. The high-level answer is that you need maybe 50% more capital as you build through to IPO or M&A. The more important question, though, is does it need to come from the cap table? Because what folks are really worried about is getting diluted from additional rounds and that’s why they never do hardware. The math just doesn’t work if 50% more capital – it just means earlier investors own less at exit compared to software. But it turns out that most hardware successes in the past haven’t used that much equity because they were using things like working capital or off-balance sheet structures. We did an analysis about a year ago on this and you can see it in the data.
VCs got really spoiled with SaaS where there is a lot of data and nicely shaped curves that give the illusion of repeatability. Physical things, by contrast, tend to have step functions, which folks really don’t like if they’re used to nice gentle curves. So you need to do some storytelling about step functions to get people comfortable with them and implement a different discipline around smoothing cash flows. But if you break down the most valuable companies in the world, none of them are just doing software. Most tech companies have some combination of hardware and software. They’ve had to figure out how to deal with capacity constraints that are real physical things. They’ve had to deal with S curves and cost curves, which are all core to hardware. Yes, there are network effects and other things that are part of the software framework of growth, but all the most valuable tech companies in the world still have to manage physical constraints and learning curves. So it’s helpful to understand the benefits of discrete manufacturing and repeating things a lot of times so that the costs come down or the performance goes up. This isn’t news, but I think a lot of folks have forgotten that this really does underpin the way that things are made.
Investors seem to be piling into deep tech, which is very hardware-driven. Has the tide finally turned?
I think it will take a long time for more than like 50% of VC to be excited about physical things as long as you can still make good money in software. It’s funny to me because there is a huge supply and demand imbalance going on in software in terms of the ratio of deals versus investors. It’s far more competitive. The idea that you could build a top-tier fund or portfolio seems hard just based on competitive dynamics in software. We have these conversations with LPs, but most still think it’s just a cute intellectual conversation. Very few seem to actually believe in any of this.
Third Sphere is largely focused on investing in climate technology companies that can make a big impact on climate change fast. That requires these technologies to be rapidly adopted at scale—what’s the secret?
There are two basic dimensions we think about when we think about adoption, and both are about reducing constraints by reducing complexity. One way to reduce complexity in a technology is to literally have fewer parts. Things with a lot of moving parts tend to struggle, and things with fewer moving parts tend to win – a solid state is probably the ideal goal. The reason is that things with more moving parts have more failure modes—combinatorially, there are just more things that are going to fail. That doesn’t mean you can’t do a good job, but it takes longer to refine and support things with more parts.
The other type of complexity that we think about is related to sales friction: when someone buys something, how much does it need to be customized before they can use it? The idea here is that a box arrives, you take the thing out and it works the same way every single time for everyone. You push a button and off you go. Reality is usually some assembly required but the goal is to avoid soft costs, which often look like construction projects.
The reason we think about technologies in this way is because whenever you get into more complex and customized things, you need to believe that more things are true other than customer demand such as energy availability, security, or policy. One of the biggest assumptions is about the availability of capital and how the financial world underwrites things.
It’s not just that things can go faster or there’s lower friction to buy or to build when things are low complexity and low customization. It’s also the fact that the capital markets are actually optimized for either simple, low customization or they look like construction projects. There is a lot of uncertainty in between. So our view is if we are trying to get climate technologies deployed in a timely way at scale we want to remove capital risk.
Third Sphere has a very unique perspective on the role of VC in climate tech. Why isn’t VC sufficient and what alternative funding mechanisms should climate tech founders consider?
It’s the Copernicus problem in the sense that VCs shouldn’t be in the center. VCs have a role, but it’s relatively expensive money and, as it turns out, relatively unreliable. Not because VCs are bad people, but the way the game is played there’s just a very lumpy supply. A founder can hit their plan and go back to an existing investor and say, ‘Okay, I did what I said I was going to do, can I have more money? and the answer over the last 12 months has been, increasingly, no. That’s nothing malicious, but VC has a cycle. People believe different things at different times.
What I wish more folks understood about VCs is that we’re not that smart and we’re not that reliable compared to say customers, partners, or other types of investors. There are often better-aligned partners for founders. So I think the best way to think about VC is as one piece of the puzzle. VCs, for example, want to see a 10x opportunity. There’s some need to prevent downside, but generally, as an equity investor, all things being equal, the top line is all you care about. But other potential sources of capital care about other things and are often more reliable. The founders need to figure out what those other pieces of the puzzle are and there are a lot of options.
Customers, for example, may be willing to put down a downpayment on your technology that is subject to certain terms and conditions. People forget that putting a $50,000 deposit on a Tesla Roadster actually worked out fine. You couldn’t start selling Roadsters any other way. But a lot of founders get stuck here. They don’t ever think that customers would put down the deposits. It’s sad because it’s a phenomenal way to do customer discovery. Someone’s willingness to put down a deposit is a pretty clear signal that you’re building something people want and it’s sad that more people don’t do it. In fact, I think the biggest risk for founders in deep tech is that they don’t really understand their customers or they assume that their customers don’t know what they want until they ship something, which I think really underestimates their customers.
The complexity and the customization framework are also helpful for thinking about lending. There are more financial products for commodity things than there are for complex, customized things. In fact, the capital markets generally have a really hard time with things that are customized. The more something looks like a construction project the less lenders like it even if there are a lot of line items that look the same like doors and windows because they need to be put together a different way every time. Once you are selling to customers, lenders will find you more attractive. The easiest way for them to understand whether you are investable is by your customers giving you money. If you make enough margin on every unit that you sell or every service that you deliver, there really is infinite money.
I get that this can feel abstracted from the real work—the science and engineering—but I’m assuming that stuff gets done right. I don’t worry about whether we can figure out hard problems in science and engineering. There’s enough talent. What I do worry about is that we sometimes just want to throw more money and people at a problem and I think there is such a thing as giving people too much money. It can make them dumb or remove the constraints that force them to think creatively about the problems they’re solving.
Is there a risk that an over-reliance on VC as the main source of funding for climate tech actually hurts our ability to fight climate change effectively?
VCs have been fairly successful at grabbing the narrative as the prime movers in funding innovation. I don’t think that’s been historically true. It’s great storytelling, but not actually reality. I think a lot of what characterized the last period of VC is bag-holding. I tell you a story so I can offload my position and return capital to my investors, but I don’t actually give a fuck about you. Climate SPACs are the classic endgame for that. It was a money bomb that you could use to go and do whatever you wanted, and most of them didn’t work out.
So I think we’re going to have cleantech 2.0. A lot of things are going to go bust. There’ll be a few survivors, but I think people will lose a lot of money. Adaptation and resilience seem like it’s going to look a lot like defense or public safety or even cyber security investing, which is perennially quite good. I’m not worried about returns, I’m more worried about the hangover from some really bad behavior in VC. Taking money out of a company that almost certainly won’t exist a few years later feels like a scam. If the difference between a company being worth a billion dollars and zero dollars is two years, we should call that a scam.
There are very few things that came out of the last tech cycle that will be around in five years. They don’t have to because the model didn’t require that; all that was required was that someone took money to generate a return for some investors and some other investors, usually in the public markets, held the bag. It’s not a popular thing to talk about and VCs use nice language like liquidity when they do talk about it, but what we actually had was a lot of bad actors that dumped bullshit on the market to make a quick return. The way we think about climate investing at Third Sphere, by contrast, is that when we get money back, the company we invested in should be around for 20 years. That’s about the timeframe to get to the scale of impact we’re underwriting for. That’s the only way we can deliver on the types of climate impact that we want to see.