The one‑page runbook template: headings and what to put under each
Start with a single A4 or a pinned doc: short, scannable and action‑first. Put the most important items at the top so anyone can act in the first 60 seconds.
- Title + brief purpose (one sentence).
- Owners: primary, backup, and on‑call contact method (phone/Slack). Include local backup if you’re in Fareham or elsewhere on the South Coast.
- Canary checks: 1–3 quick tests (eg: create test lead → expect CRM record in 60s). Include the exact test data to use.
- Kill switch: where to pause the sync (Zapier/Make toggle, toggle in middleware, or CRM integration toggle) and who must flip it.
- Quick rollback: the minimum steps to stop damage (pause, snapshot, isolate recent batch, reenable read‑only).
- Common failure causes + one‑line fixes: auth expired → reauthenticate; schema mismatch → revert last field change; rate limit → pause and backoff.
- Validation queries: 2–3 queries or saved lists that prove correctness (eg: count of records created in last 5m, sample record values).
- Re‑sync steps: exact commands or UI clicks to retry or reprocess a batch, and where to find the logs.
- Communication script: short internal Slack message and a customer‑facing line if outage affects customers.
- Post‑mortem owner and next review date.
Build it in an afternoon: a 3‑step, practical plan and how to surface the runbook
Spend 90–180 minutes to write and publish a usable runbook that the whole team trusts.
1. 30–60 mins: Draft the template using the headings above. Populate Owners, Canary checks and Kill switch with real contacts and exact steps. Keep language command‑style: "Do this, then that."
2. 30–60 mins: Add the validation queries and re‑sync steps; test the canary checks and record the expected outputs. Note platform specifics for HubSpot, Salesforce, Marketo or Pardot, but keep the action language platform‑neutral (eg: "create test lead → wait for CRM id").
3. 30–60 mins: Publish and surface it: pin the doc in your team drive, add a Slack shortcut or saved message for "Runbook - Integration", and add a short CRM property or record (eg: Integration Runbook Link) so team members can open it from the CRM record.
Low‑tech surfacing works better than another notification: a pinned doc, a Slack shortcut and a CRM property together let anyone find the runbook during a busy shift. If your team runs workflow automation across the South Coast or needs help standardising these steps, see workflow automation support for small teams.
Make it live: set a monthly 15‑minute check for the runbook (owners confirm contact details, canary checks still pass, and any recent failures are summarised). Add it to new‑starter onboarding so knowledge doesn’t live in one head.
Two short examples and a 30–90 minute drill checklist
Example 1 — webhook → CRM sync: your form platform posts a webhook to middleware which writes to HubSpot or Salesforce. Canary: submit a test form with the string CANARY123 and confirm the CRM record appears with the same value in 60s. Kill switch: disable the webhook endpoint in the middleware; Quick rollback: pause the CRM automation that assigns owners and revert the last batch in middleware logs.
Example 2 — scheduled ad‑report import: nightly CSV pulls from Ads API into Marketo/Pardot or a spreadsheet that your CRM imports. Canary: create a synthetic campaign row with a known timestamp and verify it appears in the CRM report. Kill switch: stop the scheduled job or move the CSV to an "_quarantine" folder. Re‑sync: reprocess the saved CSV or run the import UI re‑try.
30–90 minute drill checklist (use this monthly or after any change):
- 0–10m: Announce the drill in Slack and confirm who’s primary and backup.
- 10–25m: Run the canary check and observe expected results. If it fails, follow the one‑line fix for the top cause.
- 25–40m: If failure persists, execute the kill switch and notify stakeholders using the communication script.
- 40–60m: Run the re‑sync steps, validate using the saved queries, and confirm records look correct.
- 60–75m: Re‑enable normal flow and run canary again.
- 75–90m: Quick debrief: record what happened, update the runbook (owner, fix, new test) and schedule the next review.
If you prefer someone to run an initial runbook build or a live drill for your team, Optira can help as a practical, hands‑on option rather than a theory exercise.