Most startups fail. Estimates vary, but let’s go with 4/5 new ventures don’t last more than a year. For startups, there are two ways to address this: you can decrease the chance of failing on a particular business, or you can just start more businesses.
In some ways, these strategies drive in different directions. Generally, the best way to increase the odds of success on a project is to do more research, and put more hours in. That also means you’ll spend longer on things that eventually fail. On the other hand, if your execution is all of the ‘throwaway’ variety, the chances of success plummet.
There are many ways to improve on both aspects, and I’ve tried to write about some of them elsewhere. But this time I want to focus on the interaction between the two: how can trying many ventures lead to higher success odds over time?
The book Happy Accidents, by Morton Meyers, contains many illuminating stories about medical breakthroughs. What unites all of the breakthroughs in the book is serendipity. Meyers’ notion of serendipity is the meeting of luck and preparation. From a review of Meyers’ book:
In the introduction, [Meyers] discusses the nature of serendipity in discovery and emphasizes that more than chance is involved. As Louis Pasteur once stated, “chance favors only the prepared mind.” The investigator must recognize the singular importance of a chance observation. Intuition also plays a large part, and as Peter Medawar emphasized, a true discovery is unpredictable and essentially a creative act.
Serendipity relates to startups a great deal. Consider the genesis of Slack. The founders of Slack started out by trying to make a massively multiplayer online game. That project went nowhere, but they spun off one of their internal tools into a product, which became Slack.
There are dozens of stories like this, where a startup team went down one path and stumbled across something better on the way.
I’ve had this experience myself, as a graduate student. My first dissertation idea was about feature deletion (i.e., how to create an economic model of consumer demand responses to the removal of a feature on a product), but I wound up writing a paper about a model of heuristic strategies for reducing choice complexity (i.e., I model how consumers arrive at a short list of products when they are faced with many options).
Let me repeat that serendipity is a meeting between luck and preparation. How can you be ready when luck joins your table? I believe that you can do this by establishing yourself, or your startup, in a certain domain. By ‘establish,’ I only mean that you should have a framework for taking advantage of opportunities. Alexander Fleming discovered penicillin, for example, because of a chance occurrence where a mold spore floated into his lab, and he saw that it kept bacteria at bay. But he had a lab. He had the means to recognize the impact of the mold. Slack had a team of software engineers and an organization built to make and sell products. I had a dataset and a bunch of training in economics and statistics to rely on. I hope the pattern is clear. For serendipity, you need a scaffold to capture and sustain insights you may encounter.
This relates to the crux of this essay: if you stick around in the same domain when you start new businesses, you get more and more established. You build a house where luck can come to visit. The alternative, jumping around from domain to domain, is like building new foundations in different places and never finishing the house.
In some sense, having an institution in a domain is like having a net with which to catch luck. One question that may arise: how wide a swathe of opportunities can an institution chase?
I don’t have a definitive answer to this question (as far as I know, nobody does), but I have some thoughts. The first thought is that it depends on your vision. Some people are able to see connections that others don’t. For example, for many years, luggage companies asked consumers for suggested improvements about suitcases. Believe it or not, nobody ever suggested adding wheels. For whatever reason, consumers did not see a connection between wheels and suitcases. On the other hand, Steve Jobs had good vision for many connections: touchscreen + phone, phone + web browser, music management software + store + mp3 player, for three examples.
A second thought about scope is that some domains or institutions are better for breadth and others are better for depth. For example, I believe software is more of a breadth domain while hardware is more of a depth domain. (Think of the extreme variety of Google software products vs. the relatively more narrow focus of Apple hardware.)
In any case, having a net is better than swinging your bare hands around. Just being established in a domain means that you are better able to extract value from new discoveries.
Back to the initial point of the essay: if you have a net, each new venture is another swing, and a big chance to catch some luck.
Richard Wiseman conducted one of my all-time favorite studies. He sorted people into two groups based on whether they believed they were lucky or unlucky. Then, he had them both read identical newspapers with instructions to find some specific words inside. However, in the first few pages of the newspaper was printed, in large letters, a notice that the participants could stop reading immediately and claim a cash prize. In other words, the subjects of the study could simply stop working on the given task and get cash from the researcher. All they had to do was notice some large words in the first few pages of the newspaper. As it turns out, a large majority of those who called themselves ‘lucky’ saw the notice and claimed the cash, while the opposite held true for the self-described ‘unlucky.’
Wiseman describes ‘unlucky’ people as being too focused on their expectations to notice profitable opportunities. In an entrepreneurial context, this means tunnel vision on the original plan, even when attractive chances to pivot present themselves.
Happily for us, Wiseman gives us some ideas for how to be luckier:
These all have analogs in the world of startups. The first and second are directly transferable. I find the third to be applicable to Jeff Bezos and Elon Musk: both have insanely positive outlooks that generate hype and excitement, which in turn generates support (in the form of monetary and social investment). The support makes the bold claims about some new product seem more realistic. Then Bezos and Musk look like visionaries, or lucky, but much of their visions became real because society wanted them to. The fourth probably applies best to individual founders who fail and, rather than getting negative, feel glad they still have resources to start again.
If you’re reading this, and you’ve been working on a startup for a while that isn’t growing like you want, take a moment to ponder. Should you kill the project? Could you spin off an intermediate or internal product you’ve developed along the way? Could you pivot? It may be time to pick your luck-net back up and take another swing.
I’ve referred to the power-law distribution before. The idea, in the context of new ventures, is to find the gigantic payoff that swamps all of the previous losses. You just need one.
Serendipity is often the doorway to the grand payoff. To get it, you need a way to capture it. You need to establish yourself in a domain, like Slack did by having an organization capable of building a software product. The more things you start, the better chances you have of getting the big payoff. The less you jump around between ventures, the better your chances of leveraging past effort. (Note: this doesn’t mean you can’t succeed spectacularly by jumping around. See Codecademy for a compelling example.) Sometimes you need to stop acting like an unlucky person and kill your current project.
Here’s hoping you find the big payoff.