How do we document and hand over automations so they survive staff changes?
Give every automation an owner, a one-page description, company-owned credentials, team alerts and a runbook, so workflows keep running when people change roles or leave.
- Time needed
- 30 min
- Steps
- 6
- Level
- Beginner
In short: Give every automation a named owner, a one-page description (trigger, steps, systems, what success looks like), a list of the accounts and credentials it uses, error alerts that go to a team rather than one person, and a short runbook for the failures you already know about. Keep all of it next to the workflow, review it when the workflow changes, and test the hand-over by having someone else fix a planted problem.
Why automations break when people leave
Most automation failures after staff changes are not technical. The person who built it knew which account it ran under, which field it depended on and what to do when a vendor changed its API. When they leave, that knowledge goes with them, and the first sign is a silent failure weeks later.
Step 1: Keep an inventory
| Field | Example |
|---|---|
| Workflow name | New client onboarding |
| Owner (role and person) | Operations lead |
| Platform and location | n8n, project "Ops" |
| Trigger | Payment received in Stripe |
| Systems touched | Stripe, HubSpot, Google Drive, Slack |
| Credentials used | Service accounts, not personal logins |
| Business impact if it stops | New clients not set up |
| Last reviewed | Date |
One spreadsheet or database for every automation across Zapier, Make, n8n and Power Automate.
Step 2: Write a one-page description per workflow
- What it does in one sentence.
- The trigger and each step in plain language.
- The fields it reads and writes, and which system owns each.
- What success looks like, and how to check it.
- Known limits: rate limits, time windows, manual steps.
Put a link to this page in the workflow's description or notes field so it is found from the workflow itself.
Step 3: Move credentials off personal accounts
Connections built on an employee's personal login stop working when that account is closed. Use service accounts or shared admin accounts owned by the company, store secrets in the platform's credential store, and list them in the inventory.
Step 4: Send alerts to a team
Every workflow needs an error path that posts to a shared channel or mailbox with the workflow name, the record and the error. Our view: make the failure loud, and reliable automation must detect, escalate and recover.
Step 5: Write a runbook for known failures
| Failure | Check | Fix |
|---|---|---|
| Expired token | Connection status in the platform | Reconnect with the service account |
| Missing field | The failing record | Fill the field, re-run the execution |
| API rate limit | Error message | Wait and retry, or reduce batch size |
| Duplicate records | Idempotency key or lookup step | Merge, then fix the lookup |
Step 6: Test the hand-over
Ask someone who did not build the workflow to fix a planted problem using only the documentation. Whatever they had to ask about goes into the documentation.
FAQ
How much documentation is enough?
One page per workflow plus the inventory row. More is rarely read.
Where should documentation live?
Next to the workflow (a link in its notes) and in one shared place everyone can find.
Who should own an automation?
A role in the business that cares about the outcome, with a named person. The builder is not always the right owner.
How often should we review?
Whenever the workflow changes, and at least quarterly for business-critical workflows.
How Octacer helps
Octacer documents every workflow it builds and can audit and document automations someone else built. Our workflow best practices are a good starting point, and Engineering Capacity can provide an engineer to own them.
Next step
Create the inventory this week and fill in the owner and credentials columns first. Those two columns prevent most hand-over failures. Book a call if you want help.
Keep going
More guides
Want it running on your stack?
We implement this guide for clients, tailored to your stack, your data, and your team's workflow, and handed over documented.