The price of a tool is not the licence. It is the migration, the relearning, and the months of reduced output while habits rebuild. Almost nobody prices those three.
What a switch actually costs
- Migration. Exporting, cleaning, relinking. Usually a weekend, occasionally a month.
- Relearning. Muscle memory is real. For about six weeks I am slower at the same task.
- Rebuilding the periphery. Shortcuts, scripts, export paths, the small automations that made the old setup fast.
- Trust debt. For a while I cannot be sure the new system is complete, so I check two places.
That fourth cost is the one that quietly destroys the benefit. A half-migrated system is worse than either system alone, because attention is now split.
When switching is still right
I am not arguing for tool loyalty. I have switched, and it was correct, when at least one of these was true:
- The old tool could not do a task that now happens weekly.
- The old tool was being discontinued or was no longer maintained.
- The switch removed an entire category of maintenance work.
Notice what is not on the list: a nicer interface, a better mobile app, a more elegant data model. Those are real pleasures and bad reasons.
How I decide now
I keep a short written list of what the current tool genuinely cannot do. If a new tool only fixes things that are not on that list, the answer is no. It sounds rigid. It has saved me roughly two migrations a year.
