Find What Is Worth Automating

What you will learn

• How to score tasks so you automate the right ones first
• Why frequent and boring beats clever and rare
• The tasks you should never automate

Automation pays back in proportion to frequency times time per occurrence times error cost. A task done three times a day that takes four minutes is worth more than a monthly task that takes an hour, and both are worth more than a task you find interesting. So build the list from your actual week, not from your imagination: for three days, note every repetitive action and how long it took.

Score each candidate on four things: how often it happens, how long it takes, how boring it is, and how much a mistake would cost. High frequency, high tedium and low mistake cost is the ideal first automation - copying data between two systems, generating a standard document, sending a reminder, compiling a weekly summary. Low frequency or high mistake cost belongs later, or with a human in the loop.

Start with the boring ones deliberately. Interesting automations are usually complex, break often and teach you little; boring ones run for years. A useful rule: if you cannot explain the workflow in five steps to someone non-technical, it is not your first automation.

Some tasks should not be automated at all. Anything where the personal touch is the value - a thank-you to a first customer, an apology, a negotiation. Anything that touches money, legal commitments or irreversible actions without a human confirmation. Anything you do so rarely that maintaining the automation would cost more than the task. And anything you do not yet understand: automating a process you cannot do manually usually automates the mistakes as well.

Finish with a ranked shortlist of three candidates and pick one. The purpose of the first automation is not to save an hour; it is to learn the pattern - trigger, steps, output, check - so the second one takes a tenth of the time.

Action step

Log repetitive actions for three days with their frequency and duration. Score them on frequency, time, tedium and mistake cost, list the ones you will not automate and why, then pick one boring, high-frequency candidate as your first workflow.