How much time should I spend on a business decision?
Short answer
Match the time to how hard it is to undo. Foundational, hard-to-reverse choices, like how your team's shared drive and naming system are set up, get real thought. Reversible ones, like where the logo goes, get decided on the spot. For a big new idea, cap the test at about a month so the downside is small and the upside can run for years.
When something will have a long-term impact and is hard to change, I put more energy into covering all my bases. When it's not a big deal and easy to change, I don't spend much time on it.
What are one-way and two-way doors?
Jeff Bezos uses those terms. A two-way door is something like signing up for a gym you don't like. You cancel. It cost you 20 bucks. Make that decision quick.
A one-way door is a foundational decision. Once you make it, everything else you do is built on top of it.
| Decision | Door | How long I spend |
|---|---|---|
| Logo on the left or right | Two-way | Seconds. If I don't like it, I'll move it. |
| Signing up for a gym | Two-way | Quick. Cancel if you hate it. |
| The folder and naming system for the team drive | One-way | Real thought before building |
Our team drive was a good example. Once a whole team and every new client is built on the same folders, naming is going to be so key. We need to use the same naming system 100% of the time. That's worth slowing down for. The logo isn't.
How do I test a big new idea without a big risk?
Make it a one-month detour. It's low risk because you're only risking a month and some work, and the reward can be incredibly high.
"If it doesn't work or doesn't pay off enough, we can just bail on it. If it does pay off, what you do in the next month, what we do in the next month, could pay off for the next 20 years of our life."
Test with what already exists, too. When I wanted to find out whether greetform's offers would make money, I didn't launch a brand-new newsletter. I already had one I was writing, so I turned greetform back on there. If that works, then we go launch other locations.
What if the build is big?
Get an estimate before you start. I asked my AI how long an internal client-management app would take. At 30-plus days, it wasn't worth it right now.
I didn't throw it away. I had everything saved in a vision doc, perfectly documented, so if and when I decide to pursue it, an agent can pick it up and start going immediately.
How do I pressure-test whether I should do it at all?
I ask AI for an honest, big-picture assessment before we go further. How would I use it? How big can it be? What would I have to do to get there? And the important one: am I actually capable of doing it, or is this one of those things that makes sense in theory but I won't do because I have too much going on? Compare it to everything else on the list, not in isolation.
What if the business suddenly has momentum?
Go. In May 2025, people wanted to give us a ton of money. I'd never had momentum like that. Turning them down would have stopped growth just like failing would, so what did we have to lose?
For me, doing three newsletters is about the same amount of work as doing 30. I'm working 10 hours a day either way, so why not do the bigger one? The main thing that holds people back is fear.
Plenty of people hear about the local newsletter opportunity and then wait years because they overthink it. I went hard and fast to 18 newsletters to see what would happen. That got me on the radar of a consultant who then wanted to help sell my software. Each action feeds the next one.
For choosing between ideas, see how I decide which business idea to pursue, and for testing cheaply, why I launch before the full plan.
Answered by TJ Larkin. From team and partner calls in 2025-2026 and my own notes to my AI agents. Last updated .