Recently, for various reasons, the pace of the project slowed down a bit, and as a result, I found my confidence in approaching the product had dropped sharply. At that moment, I instinctively knew. If I keep going like this, the project will go nowhere.

In the past, I’ve always argued that “projects need to proceed quickly and with momentum.” But in most cases, it wasn’t actually done that way. Setting aside how to solve it (that’ll probably be the next article), I thought deeply about why it happens that way. And the conclusion I reached was, “The confidence graph doesn’t go up-and-to-the-right as much as we think.”

The actual graph on the right—doesn’t it look familiar? Yes, it’s the Dunning-Kruger curve (a curve showing that the level of knowledge and level of confidence aren’t that proportional). And if you don’t release the project quickly and get feedback, you naturally face the curve of declining confidence. At that moment, you now fall into the “curse of quality.”

The curse of quality is one of the common curses in startups, where you hustle and bustle around and put a lot of effort into finding “good quality.” Of course, seeking quality itself isn’t a bad act, but it slows down the product’s feedback cycle and ultimately causes the product to lose momentum.

The moment a product loses momentum, whether to proceed with the project is determined entirely by the weighing of sunk cost versus expected utility. If the expected utility is very high and you must continue, you can seek ways to bring back the product’s momentum, but for side projects, this is the point where things fizzle out. (Or the same would be true for startups trying to find initial PMF.)

That’s why you need to think about ways to extremely reduce the feedback cycle.