$345 million per employee

In 2014, Facebook had a problem. They were being out-competed by WhatsApp, a 55-person startup. Facebook reacted by buying WhatsApp for $19 billion. How could a business be worth $345 million per employee? A century ago, it would have been inconceivable that a 55-person team could out-compete Standard Oil's 60,000 employees. Industrial businesses grow with labor and distribution. Digital products need upfront investment to build them but can deliver lasting value with low marginal costs. WhatsApp leveraged the software business model's attributes to generate such an efficient business.

Software has no intrinsic value. It's just a long number. So, to be worth anything, creators must solve a problem. WhatsApp addressed human psychological needs by building a communication tool and did a better job than the existing SMS system. While Apple's iMessage made SMS obsolete for iPhone users, the rest of the world ran on Android and lacked this alternative. One winner tends to dominate the market in software. In this case, people chose WhatsApp over the thousand other messaging apps created around the same time.

The upfront costs of building software can be enormous. WhatsApp managed these costs by building a focused product that did one thing well - messaging. More complex products take more resources to make them. For example, Facebook spent $1 billion to buy Instagram in 2012. At the time, Instagram had no revenue, so Facebook spent over a year and a half building an ad network on top of their billion-dollar purchase. It was only after this significant investment that Instagram finally started to make money. But, that investment has paid off - today, Instagram earns Facebook $1 billion every three weeks.

Zero-dollar marginal costs define the software business model. When a restaurant sells a steak dinner for $100, they spend about $90 fulfilling that meal's costs - ingredients, labor, and more. When Microsoft sells a license of Windows for $100, it costs almost nothing to fulfill the download. Low marginal costs enabled WhatsApp's 34 engineers support 450 million users - more than the USA population. They wrote the code once, then can distribute it to an almost unlimited number of devices.1

Media offers the closest parallel to software's business model. An author puts time and effort into writing one book; then, they can sell that book to millions of people. When people pay $10 to download the e-book, it costs nothing extra to fulfill. In extremes, authors can earn billions of dollars. Yet, the value of media is ephemeral - nobody wants to buy last month’s newspaper, and they certainly won't keep paying for a book after they've read it. So, media businesses must continue to create new content to maintain revenue.

Software's ability to provide ongoing value differentiates it from media. When people buy software, they want a tool that solves a problem. As long as the software solves that problem, they will continue to pay for access (or updates). This ongoing value creates enduring revenue. WhatsApp cost users $1 per year. Though that sounds cheap, multiply $1 by 450 million people, and they made about $8 million per employee annually.

While software is expensive to build, it's cheap to run and creates constant value. WhatsApp leveraged the best parts of the software business model to build a business worth $345 million per employee. They made a simple product that people wanted. Next, they leveraged low distribution costs to grow to hundreds of millions of users. Finally, they earned a massive stream of recurring revenue by setting up a subscription. While WhatsApp was an outlier, it demonstrates a fundamental economic shift. In the Information Age, small groups of people can build big businesses.

  1. I'm simplifying - WhatsApp also had to run servers to relay messages between devices. But, their servers didn't store much data - which kept costs low.

Subscribe to future posts

An occassional tech newsletter
by Philip I. Thomas