Org health monitoring is the difference between an admin who reacts to fires and one who prevents them. A practical monitoring stack:
Daily / real-time:
- Failed Flow Interviews — Setup -> Process Automation -> Paused and Failed Flow Interviews. Check daily; investigate failures.
- Flow / Apex error emails — make sure they go to a monitored alias, not a person who's on holiday.
- Login History — quick scan for unusual login patterns, failed login spikes.
- Integration error logs (Mulesoft, custom) — wherever inbound/outbound API errors land.
Weekly:
- API Usage — Setup -> System Overview. Check 24-hour API usage trend; flag if approaching 80%.
- Storage — data + file storage utilisation.
- Background Jobs — long-running or stuck async jobs.
- Apex Test Coverage — run all tests; if coverage dropped below 75% you can't deploy.
- Setup Audit Trail — sweep last week's metadata changes; spot anything that wasn't through proper change control.
Monthly:
- Health Check score — re-run, fix any new findings.
- Permission audits — who has View All Data, Modify All Data, "Manage Users", "Customize Application", "Author Apex"? Review for least-privilege.
- Sharing rule audit — total rule count per object (sharing rules contribute to recalc time).
- Inactive users — deactivate users who haven't logged in for 90 days.
- Unused custom fields — Field Trip or similar tool to find fields that are never populated.
- Stale dashboards/reports — last accessed > 6 months; archive.
Quarterly:
- Field-Level Security audit — sensitive fields (PII, financial, HR) — confirm only the right profiles/permission sets have access.
- Record-level security audit — sample 5 users per role, verify they see what they should and not what they shouldn't.
- Object usage audit — record counts, growth rates, indexes; identify objects approaching LDV.
- Connected Apps audit — Setup -> Connected Apps OAuth Usage. Disable unused; renew certificates due.
- Code complexity — identify Apex classes/triggers/flows that are hard to maintain; technical debt list.
Tools that help:
- Salesforce Optimizer — built-in tool that runs a comprehensive analysis on demand.
- Event Monitoring (Shield) — continuous detail logs for serious orgs.
- DevOps platforms (Gearset, Salto, Copado) — change tracking, audit, code analysis.
- Salesforce Inspector / Apsona — admin power tools for deeper inspection.
Establishing a cadence: pick 4-6 things to check daily, 6-10 weekly, 8-12 monthly, fewer quarterly. Document the runbook. Train a backup admin to do them when you're out.
The pattern that distinguishes mature orgs: monitoring is in the calendar, not just on a checklist that gets done when something breaks.