Choosing a Problem You Can Stay With for Five Years

Founders pick problems for many reasons in the early days, including whatever feels novel, whatever they can pitch easily, whatever a recent article made them think about, or whatever a friend mentioned over coffee, and the result is often a company that runs out of energy somewhere between month nine and month eighteen because the founder has lost interest in the problem long before the market has decided what to do with it. The longer truth is that almost any startup worth building will need five years before the work pays off in the way the founder hoped it would, and the question of whether you can stay with the problem for those years matters more than whether the problem is the cleverest one you could have chosen.

This piece is about how to pick for endurance rather than novelty, and it is meant for founders in the earliest stage, before the company exists, when the choice you make is still cheap and the wrong choice still costs only an afternoon of thought.

"The trouble with novelty as a filter is that it stops working almost immediately once you start the work"

Why novelty is the wrong filter

The instinct to chase novel problems is reasonable, because novel problems feel important, they make for better introductions, and they reassure you that you are not late to whatever the market is doing. The trouble with novelty as a filter is that it stops working almost immediately once you start the work, because what felt novel during a brainstorm rarely feels novel during the third month of building, and the parts of a startup that take the longest are also the least novel, including customer support, billing, hiring, slow product iteration, and the patient explanation of what you do to people who have never heard of it.

Founders who pick on novelty often discover, somewhere in the second year, that the novel idea was the appetizer and the unsexy execution was the meal, and many quit at that point because the work they signed up for has stopped resembling the work they imagined.

The test that actually predicts endurance

A more useful test is to ask whether you would still care about the problem on a Wednesday afternoon in the second year of the company, when nothing is going particularly well, the inbox is full of small problems, and no one outside the team is paying much attention to what you are building. The Wednesday afternoon test is unromantic on purpose, because the choice that matters is not whether you can be excited about the problem on a good day but whether you can do honest work on it on a forgettable one.

Most founders, when they apply this test honestly, find that their list of candidate problems shrinks significantly, because many of the ideas that felt exciting on a Sunday evening look thin when you imagine living inside them for the long stretch of unglamorous middle years. That shrinkage is useful information, and it is far cheaper to get it now than to discover it twenty months in.

Three deeper signals to look for

There are three deeper signals that suggest a problem is one you can stay with, and any one of them on its own is a weaker indicator than the combination.

1. The first signal is that you have some personal connection to the problem that did not start with thinking about a startup. The connection can be that you experienced the problem yourself, that someone close to you experiences it, or that you have spent enough time around the problem in other contexts to find it interesting before you knew there was a company in it. This kind of connection is not strictly required to build a good company, but it does make the boring middle years bearable in a way that pure intellectual curiosity does not.

2. The second signal is that the problem keeps surfacing in your reading and conversations even when you are not trying to think about your company. The problems that stay with you outside of work hours are the ones that have a chance of holding your attention across years, while the problems that only show up when you sit down to write a pitch tend to fade as quickly as they appeared.

3. The third signal is that the people who experience the problem speak about it with quiet frustration rather than enthusiasm. Problems that get loud praise in surveys are often problems that people enjoy complaining about but are not desperate to solve, while problems that get quiet, almost resigned descriptions are often the ones that have shaped someone daily life enough that a real solution would change something for them. Listening for the difference in tone between those two reactions is one of the more underrated skills an early founder can develop.

What to do if you cannot pass the test

If none of your current candidate problems pass the Wednesday afternoon test, that is not a failure, it is simply a signal that the choosing is not done. Take another month, and instead of brainstorming, spend the time inside the lives of people whose work or routines feel adjacent to areas you find interesting.

Talk to them, observe what frustrates them, and write down the problems that come up more than once. The list you build this way tends to contain better candidates than any list produced by sitting alone with a notebook, because the problems that actually matter to people rarely arrive fully formed in your own imagination.

If you find yourself drawn to a problem after spending real time with the people who have it, that is a strong indicator that you have something to stay with. If you cannot find a problem you want to spend time on after a month of looking, that is also a useful finding, because it suggests the next right move is to wait rather than to commit to something that will not hold you.

A closing thought

The earliest choice a founder makes is which problem to spend years on, and although the choice looks lightweight when it is being made, it has more downstream consequences than almost any other decision in the first year of the company. Pick for endurance rather than novelty, look for the quiet signals rather than the loud ones, and trust the Wednesday afternoon test more than the Sunday evening one, because the founders who are still building four years from now are almost never the ones who picked the cleverest problem at the start, they are the ones who picked the problem they could still face on a quiet day in a difficult month.