Sometimes it is better to lose a great deal than risk making a terrible one.
Early in my career I thought the senior people in the room knew the answers. The longer I did the work, the more I understood that they mostly knew how to decide when there were no answers. That is a different skill, and almost nobody teaches it.
The technology choices that matter are rarely the ones with a clean right answer. Should we rewrite this or live with it another year? Buy the platform or build the thin version ourselves? Hire now or wait two quarters? You can gather data for a week and still be looking at a fork where both paths are defensible. At some point the spreadsheet runs out and a person has to choose. That moment — choosing well without certainty — is most of the job.
I have a line I keep coming back to: sometimes it is better to lose a great deal than risk making a terrible one. It sounds cautious. It is really about knowing which mistakes you can recover from and which ones you cannot. A lot of judgment is just being honest about that difference before you commit, instead of after.
When I come into a company, founders sometimes expect me to walk in and announce the decision. I try not to do that, even when I have a strong opinion. If I hand a team my answer, they get the answer but not the reasoning, and the next hard choice lands back on my desk. So I do the slower thing. I lay out the options plainly. I name what we know, what we are guessing, and what we would have to be wrong about for each path to fail. I say what I would do and why. Then I hand the decision back to the people who have to live with it.
That last part matters more than it looks. The team has context I will never fully have — the customer who keeps asking for the thing, the part of the codebase everyone quietly dreads, the deadline that is real versus the one that is just loud. Good judgment is local. My job is to bring structure and experience to the conversation, not to override the people closest to the work.
It also keeps me honest about what a fractional CTO is for. I am there to raise the quality of the decisions, then to make myself less necessary. If a company needs me in the room for every call a year from now, I have done something wrong. The goal is a team that decides better on its own, with a clearer sense of which bets are reversible and which are not.
None of this is glamorous. There is no framework that turns uncertainty into certainty, and I distrust anyone who sells one. What you can build is a habit: name the real trade-off, be honest about what you do not know, choose the path you can recover from, and move. Do that a few hundred times and people start calling it instinct. It is not. It is just judgment, practised out loud.