After I posted last week, a number of better articulated posts emerged on the matter as well (1, 2). The common theme in these posts was that margins matter. It looked to me that some interpreted this as “you can’t build great hardware companies”, and then responded “but look at Apple”.

Obviously, this was as a misinterpretation - building software (where new products have a marginal cost of zero) is one way of getting margins, but that alone is not enough either. You need a moat - something that allow you to capture part of the value created. But you can capture value with both software and hardware companies. Since we’ve invested in a bunch of hardware companies, I wanted to give my top two reasons why building a hardware company can be a good idea (despite a COGS > zero).

The first one is willingness to pay. You’ve probably seen the memes before; how we’re unwilling to part with $2 for a mobile app doing something useful while in the queue at Starbucks for a $7 coffee. You could call it a mental flaw or cognitive bias, either way it’s working in favor of those building “atoms”.

The other one is that products passively market themselves. In order for an app to get widespread organic growth, it needs an NPS score that is through the roof. We rarely look at other people’s phones/computers, so discovery comes through others actively choosing to talk about software they like. It’s different with hardware. People learn passively about your iPhone, AirPods, car and e-scooter when you use these products - without you having to broadcast it. Meaning that when you have a product people use, it can easily spread like crazy.

There are a lot of reasons not to build hardware. Cash flow complications, fewer iteration cycles (you can’t patch a “hardware-bug” that easily once you’ve shipped), each new product has a marginal cost and so on. But that doesn’t mean you can’t build great hardware companies. Be cautious with the associated uncertainties, and take advantage of the obvious strengths.