Choosing technology without regretting it in two years
Four considerations
Who will maintain it. Technology nobody on the team knows adds months of learning and a dependence on one person.
Maturity. How long it's existed, who's behind it, and how easy it is to find answers to real problems.
Fit. Whether it was made for your kind of problem, or you're bending it.
Exit. How much it will cost to replace in two years.
An innovation budget
On each project you can afford one new thing — two at most. The rest should be familiar and boring, or every incident becomes an investigation in unfamiliar terrain.
How to really check
A two-day prototype on the hardest part, not on the demo in the docs. And look for last year's reports from whoever ran it in production — that's where the truth is.
Going deeper
Write a short decision document for every significant choice: what was considered, what was chosen, and what would make us reconsider. The last section is the important one — it turns a decision from a fixed opinion into a condition you can check in a year.