Skip to content
All insights

The inherited automation checklist: 12 checks before you touch a workflow someone else built

By Richard Mabbun, Founder and Principal, 1AltX ·

Inherited automations are one of the most common reasons businesses call us. A contractor built a set of Zapier, Make or n8n workflows, or wired up the CRM, and then moved on. Now something is failing, nobody knows why, and the tempting move is to open the workflow and start changing things.

Do not start there. A workflow someone else built usually carries assumptions you cannot see from the editor. These twelve checks are what we run, read-only, before we change anything. You can run most of them yourself.

Why inherited workflows break

Automations rarely break because the platform failed. They break because something around them changed. A field was renamed in the CRM. A connected app updated its interface. The person whose login powered the connection left. A new sales process created records the original builder never planned for. The workflow did exactly what it was told, and what it was told no longer matches how the business works.

That is why the first job is understanding, not fixing. Every change you make without the full picture risks fixing the visible symptom while breaking something downstream that nobody is watching. The checks below are grouped into four areas: ownership and access, what the workflows actually do, how healthy they are, and what they cost and risk. Work through them in order and keep notes as you go. The notes become the documentation the original builder never wrote.

Ownership and access

  • 1. Who owns the accounts? Confirm the automation platform, the CRM and every connected app are in accounts your business controls, not a former contractor's personal login. If they are not, fix that first.
  • 2. Which credentials does each connection use? Many inherited workflows run on one person's credentials. When that person leaves or changes a password, the workflow fails quietly. List every connection and whose identity it uses.
  • 3. Is there any documentation at all? Look for a handover document, comments inside the workflows, or a shared folder. Even a partial description saves hours.

What it actually does

  • 4. Map every trigger. Write down what starts each workflow: a form, a new CRM record, a schedule, a webhook from another system. Webhooks are easy to miss because they are configured in the other system.
  • 5. Map every write. List every place the workflow creates, updates or deletes data. Writes are where a careless change does damage, so you want the full list before you edit anything.
  • 6. Find the hidden dependencies. Look for workflows that call other workflows, lookup tables in spreadsheets, custom fields the workflow relies on, and filters that silently skip records. Renaming a field in the CRM can break a workflow that nobody remembers exists.

Health and failure

  • 7. Read the run history. Most platforms keep a log of recent runs. Look at the failures, but also at runs that succeeded with odd results, such as records created with empty fields.
  • 8. Check what happens on failure. Does anyone get notified? Does the workflow retry, stop or carry on with partial data? Many inherited builds fail silently, which is why problems are found by customers or the finance team instead of by an alert.
  • 9. Look for duplicates and loops. Two workflows reacting to the same event, or a workflow that updates a record which triggers itself again, can create duplicate records and burn through task limits.

Cost, limits and risk

  • 10. Check plan limits and usage. Task, operation and API limits differ by platform and plan. A workflow that runs far more often than it needs to can push you into a higher tier or start failing when limits are hit.
  • 11. Check where sensitive data goes. Note any workflow that moves client, financial or personal information, and which third-party services it passes through. If you are a regulated firm, this is the list your compliance team will want to see.
  • 12. Decide keep, fix or rebuild for each workflow. With the facts above, sort every workflow into one of three groups. Some are fine and only need monitoring. Some need a targeted fix. Some are so tangled that rebuilding properly is cheaper than patching.

How to make the first change safely

Once the checklist is done, make changes the boring way. Duplicate the workflow, test the copy against sample data, turn on failure notifications, and only then swap it in. Change one thing at a time, and write down what you changed and why. The next person who inherits the system, including you in six months, will be grateful.

If the checklist turns up more than you have time for, that is what our systems review is for. We run these checks read-only, write up what is broken and why, and give you a fixed-price fix list. Nothing in production changes without your approval, and you can have us fix it or hand the plan to whoever you choose.

Tell us what is not working.

A system that should talk to another one and does not. An automation nobody can explain. A technology decision you would rather not make alone. Book 30 minutes and you will leave with a straight answer and a next step.

Book a 30-minute call

No cost. No obligation. You talk to the person who leads the work.

What happens on the call

  • We listen to what is not working.
  • We tell you honestly whether we are the right fit.
  • You leave with a next step, and if useful, a written scope within two business days.