I’ve been thinking a lot about high-failure strategies. Strangely, I think this may be the single biggest shift in mindset that I’ve experienced over the last 5-7 years. It’s somewhat related to the idea of power-law distributions, but is not exactly the same thing.

Here’s the idea. If you are facing a problem, there are a few different ways to tackle it. One way to frame these options is by likelihood of success. Are you taking a high-probability approach or a low-probability approach? As an example, let’s say that the problem at hand is trying to find a job. A high-probability approach is to talk to companies in which your friends work and are actually in positions to influence a hiring decision. A low probability approach is to cold email the CEO and try to get a job.

These approaches can also be repeated over time. What’s interesting is that some approaches, when repeated over time can yield different end results. It’s helpful to think of it in terms of a handy 2×2 matrix.

So, on the vertical axis is that chance that a strategy exists at any one time. On the bottom axis is the chance that a strategy succeeds if repeated over time. There are four handy quadrants.

The bottom left is an easy one. Chances of success are low no-matter what. You should not attempt this strategy, because you are bound of fail.

The top left strategy is very dangerous. These are situations where a strategy may work out well if employed once or twice. But if repeated over time, the odds are not in your favor. The examples here are pretty obvious. Doing something illegal. Taking drugs. Drinking and driving. But some are less obvious, like being a jerk to win a deal. In a multi-turn game, you have to look beyond the outcome of a single turn.

The top right strategy is where we tend to focus our time. We tend to like strategies that have a pretty high probability of success at each turn, and have a pretty high probability of success if repeated consistently. It’s also what we are trained to hone in on. In school, we are trained to show steady and consistent progress towards a goal. Training for a sport or a musical instrument leads to improved skill over time. It’s usually not an exercise in complete futility along the way.

The problem with the top right strategy is that everyone will gravitate towards these. So it’s really tough to find an approach that someone else isn’t employing, so you have to be that much better (or lucky) to win.

The bottom right strategy is fascinating because it leads to great outcomes, but is much less competitive. It’s much less competitive because most people are not trained to find these strategies and are not trained to have the stomach to execute on these strategies for a long enough period of time. Funny enough, I often find that the better pedigreed the person, the less they are likely to gravitate towards these approaches.

Back to the job searching example. When I speak to undergrads or b-school folks about how to get a job in a great startup, I usually give them the following advice. Step 1: Make a list of 50 companies that you think would be awesome to work for. Step 2: Talk to every single one and try to get a job. Step 3: If step 2 fails, come back and talk to me.

Usually when I give this advice, I see fear in the other person’s eyes. It sounds so daunting and uncomfortable. But it works. Every time I get this advice, I never have someone come back and tell me it doesn’t work. Many times, the person doesn’t employ this strategy and does something else. But often, people do follow some version of it, and they find something great.

This is not to say that trying to reach out to direct relationships isn’t great. It’s just limiting, and limiting your option set is not likely to yield the best potential outcome. When I was in business school, I got a job at a venture backed startup from a cold email. Tristan Walker recently shared the email he sent cold to Dennis and Naveen at Foursquare that changed the trajectory of his career as well.

As I’ve reflected on this, I find that this proves to be a great strategy for a lot of challenges in life, and often, it’s best when the stakes are highest.

In the business of investing, a few firms have pursued these sorts of strategies with pretty interesting success. In particular, Nicholas Taleb (from Black Swan fame) and Summit Partners, which pioneered the cold calling sourcing model in growth equity. In fact, I’ve noticed that Summit alumni tend to be very unique problem solvers because they have the high-failure strategy ingrained in them such that the lower-right quadrant is where they love to exist.

I also find that many of the most valuable but least followed pieces of startup advice also fall into this category. Two examples in particular:

1. Talk to your customers. It’s amazing how founders all set out to be customer focused, but very quickly stop taking the time for in-depth conversations with their customers. They reason that it’s more time efficient to do other things or to gather feedback quantitatively based on surveys or user data. But talking in depth with customers all the time is insanely valuable, especially for early stage companies. There is a huge correlation between founders who do this and ultimate success of their companies. But the challenge is that any one customer interaction might not yield very much. In fact, after two or three, it seems like you are hearing the same thing over and over again. So it’s easy to move on.
2. Do things that don’t scale. Often, high-failure strategies are not terribly scalable. Also, people who have only worked at large scale businesses have a very hard time imagining non-scalable solutions. This is the advantage that early stage companies have over large businesses. They can cold call companies to try to get their first customer. They can ship broken or unsophisticated products but then fix them through manual intervention. They can change a user-flow over and over again because it doesn’t really matter if a tiny set of users have a sub-optimal experience and complain or leave. You can also blow a lot of money on marketing experiments that might seem dumb today, but will only seem more dumb down the road when the dollars and stakes are bigger.

What are other examples of high-failure strategies that you’ve found effective?