Then count the other side, which people systematically ignore: the cost of moving is much larger than the evening it takes to try.
What a switch actually costs:
Migrating what exists. Usually the largest and always underestimated.
Relearning. You are fast in the old thing and will be slow for weeks.
Everything downstream. Scripts, integrations, documentation, whatever your team knows.
Being an early adopter. Fewer answers when you get stuck, more rough edges, and the chance it is abandoned.
Against that, the benefit has to be substantial rather than pleasant. My rough test: is this at least twice as good at something I do often, or does it remove a problem I actually have? Marginally nicer is not worth a migration, however good the demo felt.
The corollary is that the best time to adopt something new is when you are starting something new. The migration cost is near zero, and the risk is contained. Trying a new tool on a fresh project and leaving the existing ones alone gets you most of the learning for a fraction of the cost, and it is what the productive constant-switchers are usually doing rather than migrating everything every month.