The staging environment that never matches production (and why)

Every team with a staging environment believes it mirrors production closely enough to catch problems before launch. Most staging environments drift from that promise within months, quietly, in ways that only become visible when something that worked in staging breaks in production.

The drift happens in predictable places: staging runs on stale seed data instead of a realistic volume and shape of real content, environment variables get updated in production during an incident and never backported to staging, and third-party integrations get stubbed out in staging for convenience and never actually exercised end-to-end.

The fix isn't more staging infrastructure, it's treating staging-production parity as something that needs active maintenance, not a one-time setup: a scheduled, sanitized data refresh from production into staging, a checklist that any production config change gets mirrored, and periodically actually testing the third-party integrations staging normally skips.

A staging environment that isn't actively maintained isn't really staging anymore, it's a different, older, quieter version of the site that happens to share a codebase, and testing against it gives false confidence that's arguably worse than having no staging environment at all.