car, computer, software
Over the course of the past few months it’s becoming clear to me that venture capitalists investing into climate technologies are now decidedly split into two major camps: Those investing into hardware, or as they often will term it, “deep tech” innovations, and those investing into software.
This is an interesting split, and you can see why each camp is doing what they’re doing.
The “software” investors have a logical mandate for focusing in on this innovation area: It’s what venture capital has traditionally succeeded at. Venture capital is very expensive capital — I like to joke that it’s the most expensive form of available capital that won’t get your kneecaps broken. So if it’s so costly, it should be deployed only where the business model can absorb such expensive capital and scale quickly enough to achieve big outcomes. In other words, capital efficiency. Invest a little bit, and hopefully get a big outcome. Software lends itself to such capital-efficient models a lot better than, say, small-scale nuclear reactor technology development and commercialization.
The problem for software-focused investors in the climate solutions sector is that most of the actual solutions, at the end of the day, are 100% tied to actual physical assets. Yes, there are some exceptions — we can all name some pure-play software-based attempts within the climate solutions sector. But most often even these efforts are still tied somehow to the transformation of our physical assets across the energy, food, water, waste and transportation sector. The sales ramp-up of a microgrid controller software solution is obviously tied to the deployment of solar + battery microgrids. The adoption of building energy efficiency controls is almost always tied to actual building controls which are of course tied to the deployment of HVAC and smart buildings hardware. Even “climate app” platforms like car share businesses, or smart metering, or commuting optimization, or trading environmental commodities (e.g., carbon credits) are all tied to the physical world one way or another, and cannot scale without hardware innovations like electric vehicles, metering hardware, and carbon-saving projects being deployed at scale.
So whenever I speak these days with software-focused climate VCs, I often hear a lament that they are finding their investment universe to be a lot more limited than they expected. Lots of cool ideas based on software, with tons of potential… but a dawning recognition that those climate-solution ideas won’t scale like software-based businesses scale in other sectors, because they’re anchored to actual physical deployments that the software business can’t tackle themselves. The microgrid controller example is a real one that I have personal, frustrating experience with — I invested into a company over a decade ago with a great software solution for that. But we learned too late that no one was going to buy microgrid controller software unless they were also buying a microgrid, and those weren’t happening quickly enough (being early looks an awful lot like being wrong). It seems like lots of climate-specific software VCs are now re-learning this lesson in real time.
The answer for such investors isn’t to shy further away from the hardware, but to lean more into it. Acknowledge and embrace that your software sales growth is intrinsically tied to hardware deployment growth. That is NOT to say that they or their software startups need to totally shift into investing into hardware themselves. But instead, a two-pronged approach of both engaging the downstream hardware project developers, and bringing them accelerated sales efforts and project capital. To use the microgrid controller example, a decade ago project capital for microgrids didn’t really exist. Credible microgrid project developers barely existed. But now both do. And so if I were an investor in a microgrid controller startup today, I would make sure they were engaging with as many of those emerging and established project developers as possible. And I would be actively introducing those developers to project capital providers and any other resources that might be important for accelerating their deployments. Software-focused climate VCs can’t ignore the hardware aspects of these industries; instead, they should be embracing their downstream hardware stakeholders and doing everything they can to help make them successful.
And the good news is that today there indeed is a “more robust capital ecosystem” that can come in alongside and amplify the software venture dollars in this way. But I don’t yet see enough climate software VCs actively engaging those other types of capital providers.
On the other hand we have the deep-tech hardware innovation VCs. Okay, quick question: Name the last successful hardware innovation in climate tech that resulted in a great venture exit without having any embedded software as a crucial part of its success. Maybe someone like a Quantumscape? I’m sure there are other examples. But it’s telling that they’re not obvious. Don’t tell me “Tesla” or such — the Model S was a breakthrough hardware innovation… that only was possible because of software innovations (battery controls, U/I, etc.). It’s basically a rolling computer and that’s one reason why early customers loved it.
Unfortunately, most often I see deep tech innovation efforts making software an afterthought. Even if it’s just firmware to make the small-scale nuclear reactor work… you still need firmware to make the small-scale nuclear reactor work.
At my firm we invest into the deployment of sub-utility scale physical assets. These deployments are only made feasible via telecommunications and remote automation software. That’s in fact exactly why these distributed and localized solutions across the energy, food, water, waste and transportation sectors are the fastest-growing segments of those markets. So we are keenly aware of how important software solutions are for the success of the physical deployments.
But too often I see the crucial areas of hardware controls, user experience, and even such prosaic areas as billing automation are addressed LAST by hardware-focused startups in our sector. There’s an assumption that they are somewhat of a commodity innovation that can just be talent acquired as necessary. Sometimes it’s even just outsourced.
To a certain extent, this is true. I can hire someone to code a lighting controls app for me tomorrow and get something that “works” by next quarter on the cheap, I’m quite sure. But as the Tesla example shows, providing the bare minimum solution is often underwhelming and buggy, but providing a great solution can lead to a breakthrough in adoption. Hardware VCs in climate tech need to be making sure that their deep tech innovators are thinking early about the software side of what they’re going to end up providing to the world, and in fact looking for ways to turn that into a competitive advantage and leverage point for accelerated adoption.
The software VC mindset is also going to end up being valuable for driving successful investment outcomes in other ways, too. After all, if you spend a billion venture capital dollars to build a billion-dollar “unicorn” hardware startup… you actually have lost money, not made it. Figuring out how to leverage non-venture dollars for the more capital-intensive aspects of the growth story will be important. And once you remove the capital-intensive aspects from requiring venture dollars (by bringing in project finance dollars, for example), what’s left other than investing into more hiring of talent and… software. So even hardware VCs in climate tech would benefit from viewing what they do through the software VC mindset.
So I personally find the current schism between “software VCs” and “deeptech VCs” in climate tech to be dysfunctional and ineffective. It didn’t used to be that way, and it’ll revert soon enough. But for right now, it’s a curious and seemingly artificial split.
