Early Stage Hardware Investing: 5X Less Competitive Than Software With Similar Returns

Hardware outcomes are similar to software, despite slightly lower survival rates and higher equity funding needs. But seed stage software deals are 5x more competitive as measured by number of deals vs number of investors in each category. Even if hardware is 2x as hard, the relatively much easier access to great teams points to a clear path to better returns from hardware.?

How do we know all of this? We’ve had our own data for a while. But this time, we asked claude.ai, Bard, and Chat GPT, and ran some Crunchbase queries. Here’s what we found:

50-80% of the most valuable companies in the world are hardware companies

Early stage is a great time to bet on hardware

Seed stage hardware investing is 5x to 16x less competitive than software investing

Hardware and pure software are just as likely to create fund returning outcomes

Hardware companies raise 20% and 50% more equity over the same number of rounds

Hardware is 10% harder as measured by survival rates

Hardware is ? of climate investing. That may go up with more focus on adaptation.

Building are great hardware portfolio today is easier than software

50% of the most valuable tech companies in the world are hardware companies

Pure software companies make up just under 50% of the top 50 tech companies in the world right now. In the top 50 companies across all sectors, that number drops. Only 22% are software only. At this point, all firms are using software in some way, just as they will embrace AI, but they’re making and using physical products too. Here's the data.?

As investors, we’d like exposure to 100% of the best possible tech outcomes. Choosing to ignore hardware would cut our exposure by 50-80%.?

Early stage is the best time to invest in hardware

The main change in growth stage VC over the last 10 years is that it’s never been easier to get signals as a proxy for traction: website traffic, job postings, updates shared from GPs to LPs who also do direct investments. For hardware there are also pre-orders, distribution announcements, partners announcements etc. Early stage is different. There is much less signal.?

Are growth stage investors doing a better job pricing risk? Looking at the Q2 2023 data from CBInsights, it seems like growth stage investors are having to write down a lot more than seed stage investors (compare series A vs series C for example). Maybe they’re doing a better job pushing down prices and getting liquidation preferences now. Still, it’s very unclear you can make a claim that seed is somehow mispriced.?

Seed stage hardware investing is 5x to 16x less competitive than software investing

Far fewer investors are competing for seed stage hardware deals. We compared NFX data on areas of investor focus at seed stage versus the number of companies founded in the last 2 years from crunchbase data in the same categories.?

Fintech has nearly 5x investors for every company founded (7330 investors; 1518 startups).

The SaaS ratio is 2.8x (4600 investors; 1661 startups).

AI ratio is 1.7x (4000 investors; 2358 startups).

The hardware ratio is 0.3x (503 investors; 1616 startups).

This difference certainly doesn’t represent a difference in risk, but much more likely this is about a relative lack of experience with hardware and at least a decade of “hardware is hard” narrative from LPs and GPs.?

Does a lower number of investors mean hardware companies can't get funding as easily? Yes, but hardware startups can—and usually do—look beyond VC for their capital needs. And the survival rate isn’t far enough apart to suggest funding is a material issue. In the end, outcomes are very similar.?

Hardware and pure software are just as likely to create fund returning outcomes?

It’s helpful to look at the top performers across the hardware vs software exit spectrum.?

Data source

Both hardware and software can produce 10x+ returns for early investors. Time to exit does not correlate strongly with returns. Software may require less capital, but hardware exit multiples can be just as high.

Hardware companies raise 20% to 50% more equity and they use non-dilutive capital

Capital intensity is perhaps the most common complaint about hardware.?

The average number of funding rounds is the same. But equity raised is between 20% of 50% more for hardware. But we think a large source of confusion comes from mixing dilutive and non-dilutive capital (grants, loans, off balance sheet SPVs). In this case, hardware capital raised is nearly 2x.?

Needing non-dilutive capital increases funding risks but that shows up in the next section on survival rates. We think non-dilutive is a another unique hardware investment opportunity and so we also built a credit fund on the Third Sphere platform.?

To try and filter down to teams that have a shot at an exit, we can filter down to only companies that raised in the last 18 months.?

We used variations of this query.

Hardware is 10% harder as measured by survival rates

The survival rate for hardware startups is about 77% compared to about 84% for software. It’s unclear how to close this gap, but one notable pattern is that a lot of successful hardware companies ship products in the first few years.?

Fitbit was founded in 2007 and had pre-orders in 2008. Sunrun was also founded in 2007 and in 2008 has a few 100 customers for their residential solar leases. Tesla was founded in 2003, but was accepting pre-orders in 2006. NVIDIA was founded in 1993 and shipped its first graphics cards in 1995. See a pattern? Revenue should not require 5 years, a new plant and multiple rounds of finance.?

In contrast, over the last few years, we have often heard folks talk about FOAK (First Of A Kind) plants. These typically mean that the company is going to take 3-5 years before they have commercial agreements, let alone revenue. We expect the failure rate of these companies to be highest because there isn’t really a clear funding solution to this company building strategy. Other things can go wrong, but getting to ship around series A, at the latest, is a good way to avoid failure.?

FOAK is also what later stage funds generally flag when expressing reservations about showing up as follow on capital. Our observation is that the best companies have the following qualities:

1. They appeal to generalists because they’ve proven unit economics and customer demand
2. They have access to other types of capital which, even in a high interest rate environment, is cheaper than dilutive capital

Hardware is ? of climate investing. That may go up with more focus on adaptation?

Generalist firms like USV and YC are building climate portfolios that have turned out to be mostly hardware. It’s true that their general portfolios are nearly all software companies (over 90% in the case of USV). But USV’s climate portfolio is at least 60% hardware, and YC’s climate portfolio is at least 70%. Third Sphere focuses only on climate, and our portfolio is about 70% hardware.?

YC’s and USV’s investments include companies building fusion reactors, fission reactors, batteries, hydrogen aircraft, carbon removal, carbon capture, mining robots, electric ships, EV charging networks, solar developers, vegan burgers, milkless chocolate, insect farms, mushroom farms and feed to reduce methane from cows. Not all hardware companies are the same (see our Climate Startup Playbook).??

Hardware startups are harder but it’s much, much easier to invest in the best hardware teams?

Hardware outcomes are already similar to software. And risk at seed stage is priced as well or better than at growth stages where more investors have access to private market signals. Hardware companies survival rates are a bit lower and most of the capital raised difference is attributed to non-dilutive capital sources like grants or loans. But the biggest difference for investors is that hardware deals are at least 5x LESS competitive.