Are we over-optimising FinOps? Random thoughts on Monday.

In FinOps, we love to chase waste, and destroy it! But at what point does optimisation itself become waste?

Software development is still more an art than a science, and the artists like it like that. For interesting projects, when you build something new, you need to try, even in prod, because focusing on cost, or scale, at that time is too early.

Should we get to something that works then make it better?

Should Shifting left be dependant on the application maturity?

Note: in Cloud you need to put serious guardrails. Why not throttling (like cloud vendors do) up to turning off your application if the cost is too high. Better offline than broke; running out of cash is the only mistake you cannot recover.

Sometimes, chasing efficiency too soon burns more energy than it saves.