A platform runs several self-managed stateful systems. What evidence should decide whether to move to managed services?

The evidence

1. What self-managing actually costs today, counted honestly: engineer time on upgrades, backups, failover testing, capacity work and incident response, plus the on-call burden. Most teams have never measured it, and the largest component is usually the opportunity cost of the most experienced engineers.

2. How good the current operation genuinely is. If failover is untested, backups are unverified and upgrades are deferred because they are frightening, managed is both cheaper and materially safer. If the team runs a genuinely excellent operation with rehearsed recovery, the comparison is much closer.

3. Whether the workload needs anything the managed service lacks — a specific extension, version, or configuration parameter. This is a hard blocker rather than a trade-off, and it must be checked first.

4. Whether the workload is near a service ceiling. Connections, storage, IOPS, replica count. Adopting a service you will outgrow in a year means paying migration costs twice.

5. What the freed engineering time would be spent on. If the answer is "nothing specific", the premium is hard to justify — and that is a legitimate finding rather than a failure of analysis.

What managed genuinely buys

Automated failover, backups and patching tested at a scale no single company matches; point-in-time recovery that works; a class of 3 a.m. work eliminated; and engineers redeployed from keeping systems alive to work that differentiates the product.

What it costs

A price premium; loss of control at the margins; provider-defined maintenance windows meaning involuntary failover at a time you did not choose; a ceiling whose breach is a migration; and lock-in proportional to how many provider-specific features get adopted.

The judgement being tested

Whether you can resist both reflexes. "Always managed" ignores real constraints; "we can run it better ourselves" is usually asserted by teams who have not measured what running it costs them.

And a specific caution: consolidating several self-managed systems is frequently worth more than the managed-versus-self question — every additional stateful system is a new place data can disagree, plus backup, failover, monitoring, patching and on-call expertise.