Make Hardware Great Again
Hardware is back in fashion.
Climate targets won’t be achieved with carbon accounting software or ESG data software alone. Most in the venture space would agree. But few have become comfortable enough with the elephant in the room, hardware-first businesses.
The purpose of this essay is to democratize the idea that deep tech is a viable venture strategy for asset managers. It is a continuation of my series on hardware companies, following my tour of El Segundo factories last month.
Venture capital’s genesis found its ‘Adam and Eve’ moment in the story of Fairchild’s traitorous 8, a hardware company leading to a plethora of offsprings (Intel, AMD, National Semiconductors…). Back in the day, venture capital funded hardware, which came out from Bell Labs or Xerox PARC. The love for software businesses is only a short term deviation in the broader story arc of Schumpeterian creative destruction, a surge of innovation reaching its maturity phase.
While software businesses present fantastic business models with quasi-nil incremental costs to revenue (a great oversimplification given rising acquisition costs), they make the uneasy sacrifice of long term MOATs due to the ease of entry / lack of defensible mechanisms beyond distribution.
Software used to have upfront barriers to entry (access to computers and servers, talent, limited playbooks), but the end of cloud subsidies, popularization of software ventures and emergence of gen AI is dropping these barriers even lower. With more competitors accessing venture dollars and achieving scale, returns are getting competed away. And as Peter Thiel said, “competition is for losers”.
Asset allocators look for durable competitive advantages, leading to blue ocean markets, and positive pricing dynamics, and we tend to find them where the barriers to entry can be built and where businesses can achieve strong staying power. Rather than seeing hardware as inherently more risky, we should reframe our assessment to see the risk as distributed differently alongside the path to scale.
Hardware / Deeptech startups typically require significant inputs to work: A core technical innovation, the right talent, scaling production, upfront investment, and demand elasticity on the client side to name a few. I’ve found that across each of these elements, the prospects for deep tech have significantly been improved over the last few years.
Staying away from hardware investing due to its complexity and cash efficiency is largely a cope out at this point. I listed below the main narrative violations surrounding hardware as a category:
“Scientific risk is too high, and VCs don’t want to invest in PHDs chasing accolades”
The technical risk is lower than perceived: these companies build systems and assemble proven technologies to deploy them as soon as possible — not the kind of binary research plays. Don Valentine had a quote which said (paraphrasing) that the job of a VC is to invest in the “D” of R&D, not the R.
“Talent has left hardware for easier software companies over the last 10 years”
Newly successful hardware companies (SpaceX, Tesla, Nvidia…) are breeding a new generation of founders chasing hard tech success based on the newly established playbooks in the space. That being said, talent density is still lower than needed. Few domain experts exist in each subvertical, which de facto gives investorsan easy differentiating factor to see which companies are built by serious people in each industry.
“Hardware requires a lot of capital upfront, and companies can’t be valued on story forever”
Successful deeptech companies aren’t waiting 7 years to get to revenue and largely rely on systems engineering to commercialize prototypes. They find their beachhead market early and get the product to their customers early. We do not invest in companies that don’t start from the customer pain points or lived 3+ years in a revenue desert.
“Hardware business models are more difficult to scale”
There are high-margin, recurring use cases for hardware IP, alongside hybrid SaaS models. Asset level efficiencies can be achieved through subsidies to help initial commercialisation. We’re also seeing a decreasing cost for inputs (off the shelf components…) and manufacturing (digital twins, contractor manufacturing, 3D printing…).
These hardware heavy companies can be incredible businesses thanks to high operating leverage and ROIC, as well as strong baked-in defensibility. The risk is clearly front-loaded, but faster iteration speed and systems engineering can help even non-technical investors understand where the story is going.
Previous startups also serve as platforms for new hardware companies to emerge (i.e Space X reducing the cost of each kg sent to space unlocks a whole world of new space businesses with realistic unit economics).
“There is too little money chasing hardware”
On the contrary, A16Z, the largest asset aggregator in VC, just raised an ‘American Dynamism’ fund. Solo-GPs and new specialized funds are emerging in the space (Also Capital, Valhalla Ventures, Cantos, Wireframe…). Larger funds are shifting towards hard tech or have partners exclusively deploying in the vertical (Founders Fund, Khosla, Lux…). Even defense officials are turning into VC investors.
“The West is no longer the production center for global industries”
There is a conflation of several factors shifting power in the other direction: New order volumes are growing again. The US manufacturing PMI was 50.3 last month (March 2024), reaching its highest level in the last 16 months. Rare are the areas where we see bipartisan alignment, but American dynamism / reshoring production is one of them. The 2022 CHIPS act and the IRA are both showing that the US government is putting big bucks behind reshoring, which it sees as a opportunity to reinforce American economic hegemony in an increasingly polarized world. Even the US commerce secretary recently used the tagline “Make hardware sexy again”
Looking at the reasons why production systems moved away from the West towards Asia, cost-of-labor arbitrage is the central answer. Other elements to complement it were rising payrolls due to Union pressure, increasingly restrictive regulatory frameworks, and rapidly decreasing shipping costs. With the rise of robotics, we see a direct path to reducing the arbitrage to a level that will propel a shift back towards ‘Made in the West’. Recent disruptions and cost volatility in containerized shipping, coupled with geopolitical instability, have made shipping and the associated longer lead times a significant supply chain risk.
This shift to hardware investing matches our DNA more closely. JP’s skill sets was built moving atoms (in this case, perishable products across 7 different temperatures in warehouses throughout China). I worked at Ara Labs, which built products for the in-car economy (car top ads, in-car vending machines…). Hardware places an emphasis on efficiency, iteration speed and building close to the customer, with an even more important focus on priorities and rhythm. We love that.
