"Managed hosting" means different things to different vendors. Most of the time it means "we run the server." For us it means a named engineer is on the rota when your site goes down at 2 a.m.
The runbook lives in our internal docs and gets reviewed quarterly. The short version: daily backups with 30-day retention, weekly DB snapshots with 90-day retention, monthly security patching window on Tuesday evenings ET, uptime monitoring at 60-second intervals, alerting that pages a real human on critical issues. We publish the SLA targets and the actual hit rates every quarter.
This post is the full document, cleaned up and shared as-is. It's how we operate; it's also a template if you want to set up something similar for your own infrastructure.