Spot the Work Worth Automating

Why this matters

Every automation you will ever sell starts with the same act: looking at someone's workday and spotting the tasks a machine could do instead. Most people cannot do this — they see 'my job' as one blur, not as separable tasks. The pattern that separates automatable work from everything else has three parts: a predictable trigger (an email arrives, a form is submitted, it is 9am Monday), mechanical steps (copy, paste, rename, send a template), and a defined output (a row exists, a message was sent). Work that needs judgment mid-stream — 'decide if this customer deserves a refund' — fails the pattern and stays human. Learn to see the pattern and you will find billable automations in any business you walk into, because a typical office worker spends several hours a week on exactly this kind of copy-move-notify work.

Demo

Watch the pattern applied to a real example. A receptionist at a two-chair dental clinic handles appointment requests that arrive by email from the clinic's website form. Her routine, verbatim: (1) open the email, (2) copy the patient's name, phone number, and requested date into the 'New Requests' Google Sheet, (3) send a canned reply — 'Thanks! We received your request and will confirm within one business day.' — from a saved draft, (4) flag the email as done. That is trigger (a website-form email arrives), steps (copy three fields, send one template), output (a new sheet row plus a sent confirmation). Every step is mechanical: no step requires her to decide anything. She does this roughly 15 times a day.

Now contrast two tasks from the same clinic. 'When a supplier invoice arrives, save the PDF to the Invoices folder and add the amount to the tracking sheet' — trigger, steps, output, fully automatable. 'When a patient complains about a bill, decide whether to offer a discount' — the trigger is predictable, but the middle is judgment, not mechanics. You could automate the edges (log the complaint, notify the manager) but not the decision. That edge-versus-core distinction is exactly how professionals scope automation projects: automate the mechanical shell, leave the judgment to the human, and say so plainly in the proposal.

Try it yourself

1. Pick one business you know from the inside — your own job, a family business, a friend's shop — and list five recurring tasks people there do by hand.
2. For each of the five, write one line in the exact format 'Trigger: ... / Steps: ... / Output: ...' — if you cannot name the trigger, the task is not ready to automate.
3. Cross out any task where one of the steps requires a judgment call, and note in a few words which step failed the test.
4. For each survivor, estimate how many times per month it happens and write the number next to it.
5. Circle the survivor with the highest monthly count — you will price it in the next task and build a version of it in this module.

References

• Zapier — How it works
• Make Help Center
• n8n Docs — Workflows