Hard Tech’s “AWS Moment”

Unlocking the infrastructure for the future of Hard Tech

In the last 20 years, “venture capital” has become virtually synonymous with “software.” Why? Because for the last 20 years, large cloud providers (Azure, Google, Amazon) have been subsidizing software companies’ capital costs, allowing these businesses to earn venture capital returns while taking growth equity risk. In this time, venture capitalists (and their LPs) enjoyed the most sustained period of strong returns in the history of venture. But that paradigm is about to change.

To understand why, look no further than AWS’s Q3 revenue growth rate, a paltry 12% YoY compared with the prior five-year average of over 30%. Azure is doing better at 30%, but still down from the 40-50% range it achieved from 2020-22. Cloud providers can try to stem this bleeding by seeding AI companies with cloud credits for equity, but that’s a short term fix rife with challenges around long-term data center sustainability, energy consumption, and expansion pace.

When cloud demand growth does inevitably slow, what will the cloud providers do next? Raise prices. And when they raise prices, all but the most mission critical SaaS companies will see their margins compress. At risk of stating the obvious, margin compression is bad for future software-focused venture returns.

So since this blog is mostly me talking my own book, it’s time to tie it back to the implications of this for Hard Tech. Three key takeaways:

1. As future (relative) returns in software VC are likely to contract, capital allocated to the strategy should also contract. I expect more capital will flow to Hard Tech and have staked my time, money, and reputation on this reality.

2. Early-stage venture returns (target 30% Net IRR over 12 years) will only be achieved through a combination of new sector focuses (like Hard Tech) and/or new portfolio construction strategies (more concentration). Because the industry is no longer subsidized by Big Cloud, the index strategy (large number of portfolio companies) will be far less effective than the past 10+ years (ie the gap between median and mean company-level returns will increase substantially).

3. Similar to how Microsoft, Amazon, and Google subsidized startup growth in software, SpaceX, Palantir, and Anduril will effectively provide subsidies for Hard Tech through shared infrastructure, customer relationships, and most importantly, talent.

Enter Starship, the “AWS Moment” for Hard Tech. Yesterday’s successful launch showed that a multi-planetary future is near inevitable, and SpaceX is leading the way. Through their Transporter missions, they’re creating lower cost opportunities for startups to build businesses on their infrastructure, much akin to what AWS did for the last two decades of SaaS startups. This is the beginning of a new era of innovation, and days like yesterday catalyze the enthusiasm and excitement that will create orders of magnitude shifts in what’s possible over the next decade and beyond.

I can’t wait.