The recent Wework debacle seems to have opened up a bit of soul-searching within tech companies and among investors. Cynics are largely looking for ways to say, “I told you so,” with the main conclusion being we have somehow gone full circle back to the cash-burning bubble experienced back in 2000.
While reflection and soul-searching is indeed needed, I think the cynics’ conclusions will likely prove to be the wrong ones when we reflect on this period in the future.
The bubble of the 2000s, was largely a hardware bubble. It was an internet bubble, but it was also a telecom bubble. And within tech, Cisco became one of the most highly valued tech companies, a reflection of the hardware orientation of the thinking at the time. Within the internet space, the vast majority of companies largely took offline businesses (which I also group as “hardware” due to the physical nature of the businesses) and brought them online. Such as Pets.com and Webvan as well as Amazon (the “hardware” for all of this is offline logistics and inventory). Amazon survived, but Amazon was and still is very much a “hardware” business. The internet had/has the potential to make these offline businesses better, but in the grand scheme of things, the businesses still are largely hardware driven.
The bubble of the 2000s gave way to a much healthier period centered around software. Many of these businesses were enabled by the Apple App Store and Google Play Store and the smartphone revolution. Others were B2B businesses trying to take apart older, legacy software such as Microsoft Excel. While software and hardware are both “tech”, the natures of the business models and financial profiles are materially different. It is easy with hindsight (as well as foresight if we consider the future), to see how software businesses should be materially better than “hardware” businesses. Software has low marginal costs and almost infinite scalability. While tech is often viewed as a single category, the variance in quality and business model is wide.
So how did we almost go full circle? I think this has largely been a result of the rise of platforms, which re-enabled entrepreneurs and investors to try their hand at “hardware” businesses.
Over the last 10 years, many largely tech businesses have been able to transform and extend themselves from singlet vertical/service to multi-vertical/service by operating as a platform. For example, Apple going from largely a hardware (+software) business to a hardware+platform(+software) business. An iPhone straight out of the box is useful, but not nearly as useful without the millions of apps that have been enabled by the App Store platform. Amazon went from largely a “hardware” 1P e-commerce, logistics-heavy company to a platform business as it shifted towards 3P marketplace as well as transformed the logistics asset into a platform through Fulfillment by Amazon. And in parallel, the creation of AWS is likely one of the most transformative moments/platforms for the industry. Google, Facebook, and Microsoft also in their own ways took their core software/service driven business as transformed themselves into platforms. There are many more examples that are less well known but impactful in their own ways as well (e.g. Nvidia taking GPU hardware and creating an incredibly valuable graphics and data processing platform through CUDA).
The rise of platforms reduced the difficulties of doing “hardware” by abstracting away challenges such as payments, logistics, data centers, etc. Hardware businesses are still hard but became incrementally less so. I think temporarily, entrepreneurs and investors went too far and forgot that hardware businesses are still hardware businesses.
The second thing that I think is more sinister is that many “hardware” startups missed the whole idea behind the platforms. They see Amazon as highly successful despite its hardware core. And they also see Amazon going deeper into hardware through things like grocery delivery. For a little while, the lesson many start ups seemingly wanted to draw was that the hardware businesses that failed in 2000 were just too early. I don’t think that is the right lesson. Many of these hardware businesses are still unlikely to prove to be great businesses in the long run. They might be big/bigger businesses now that there are billions of internet users instead of millions. But they won’t be infinitely scalable like software. The reason why they might work now is because the platforms are doing it. Platforms can fund these lower margin ventures with their existing high margin businesses. THAT should be the lesson. Not that low margin, “hardware” businesses can become software businesses. If that ever happens, it would likely validate the other part of the equation – a hardware business would necessarily need to figure out how to transform itself into a platform like what Amazon did 15 years ago with AWS.
That is what I am implicitly betting on with Uber, the only “hardware” internet business in the paper portfolio. The most relevant businesses now and likely into the future are platforms. But there is also value in integration as demonstrated by Apple. Platforms will/may have software, services, and hardware businesses. But investors would likely do well to remember which part of the chain the value truly comes from. Wework didn’t get the memo before investors ran for the exit.