January 6, 2021
The famous and controversial tech investor Keith Rabois tweeted the following back in 2017:
Formula for startup success: Find large highly fragmented industry w low NPS; vertically integrate a solution to simplify value product.
If you're unfamiliar with Rabois, you should check out his resume. He was an early employee at PayPal and parlayed that experience into executive roles in the early days of LinkedIn and Square. As an investor, he was an early backer of Palintir, Lyft, and Airbnb. He also came up with the idea for Opendoor and cofounded the company, which led to another multibillion-dollar exit.
Believe it or not, the preceding is a fairly small sample of his overall list of successful ventures. Whether you like him and his pugnacious Twitter feed or not, he's an outlier among people trying to pick and choose winners in the startup world. A formula for startup success would sound ridiculous coming from most people, but Rabois's track record is long enough that it's worth taking a closer look at his formula and breaking it down with some examples.
"Large" here is self-explanatory. Rabois needs to identify large industries because he is a venture capitalist. The math of venture capital is such that it's not enough to achieve what "normal" investors would consider a good return, like 10% or so. A rule of thumb is that 10x is the minimum return for a good VC investment because venture capitalists need their winners to be big enough to make up for their many inevitable investments that go to zero. Rabois focuses on early-stage companies, so he is facing even higher failure rates than the typical VC, and focusing on large industries is one way to avoid mediocre winners.
The "highly fragmented" part here is perhaps more interesting. A "fragmented" industry is one where there isn't a dominant incumbent. Online search is not a fragmented industry because almost everyone goes to Google for that. Selling houses, however, is highly fragmented because there are many channels homeowners can use when they are ready to sell. One of Rabois's theses behind Opendoor is that the company will consolidate that fragmented industry by giving homeowners an obvious first choice for where to sell their home the same way internet users have Google as the obvious first place for search.
"NPS" is net promoter score. It's a standard way that tech companies measure customer satisfaction. If you've ever answered a question like, "On a scale of 1-10, how likely are you to recommend this web site to a friend or family member?" then you've contributed to a company's net promoter score. All that matters here is that Rabois is trying to find industries with low customer satisfaction and, if possible, he wants to quantify that level of dissatisfaction.
It's important to note that fragmented industries do not all have low NPS scores. For example, the Italian restaurants business is highly fragmented, but people love their local favorites. The taxi industry, by contrast, was highly fragmented and had low customer satisfaction before Uber and Lyft entered the market.
What does "vertically integrate" mean? To understand this, we need to quickly review the concept of a "value chain," which is the series of steps that delivers a product to an end user. If you're sitting in a chair, you can imagine a simplified version of the value chain that produced that product. A lumber company sourced the wood, then a manufacturer made the chair, then a retailer sold it to you. A "vertically integrated" chair company would own more than one step in that process.
In tech, Apple is the most prominent vertically integrated company. In addition to building both the hardware and software their end users purchase, they have gotten into building more of their components over the years, even designing the chips on which many of their devices run. Microsoft, by contrast, has traditionally been horizontal. The core of their business for years has been software that runs on many different types of machines, though they have more recently gotten into hardware.