A mistake I've done often as a founder is not solving problems in order.
You can think of building a startup as a time sequence of things you do. And there is an optimal order of doing things.
Let's take an example in building software.
A foolish version of myself thought that clear types and a well-structured repo are not important at day 1. The objective was to move fast and get something into the hands of customers quickly. So I took many shortcuts in building things which gave me the illusion I was moving at the fastest available pace to me.
But here's the problem: when you do this, you are paying a tax on your speed a few weeks, even days, down the line. Because as your software evolves you will spend more time on shipping a new feature or fixing a foundational bug.
The better way is to do things in order. Structuring types in the right way, setting up a clean structure, and not sacrificing on quality early on. It might make you 20% slower on the first days, but ultimately will make you >50% faster after day 3 or so from my experience.
And this is also true for company building general.
But there's also a balance to it. You don't want to solve problems today that don't exist yet.
Instead of adding a customer support platform to your product with 0 users now you should implement just a simple email you and your team gets. And then as volume becomes a true problem you can implement a better solution.
But still, doing things in order, one step at a time, is what increases the speed at which you move.