Which Task Should You Automate First? A 15-Minute Way to Decide

Ask any small business owner what they would automate if they could, and you will get a list. Invoices. Onboarding emails. The Monday report. Copying orders from one system into another. The list is never the problem. The problem is picking.

Pick the wrong task and you spend weeks building something that saves twenty minutes a month, or something so fragile it breaks every time a client fills in a form slightly differently. Pick the right one and the time it frees up pays for the next project.

This is the method we use in the first conversation with a new client. You can run it yourself with a pen and fifteen minutes.

Step 1: Write down everything that feels repetitive

Don't filter yet. Spend five minutes listing every task that makes you or your team think "I've done this exact thing before." Some prompts that help:

• What do you copy and paste between two tools?
• What do you check every morning or every Monday?
• What do new clients, orders or hires trigger?
• What would fall apart if one specific person went on holiday?

Most teams end up with 10 to 20 items. That's normal.

Step 2: Score each task on four questions

Give each task a score from 1 to 5 on these four questions.

1. How often does it happen? Once a quarter scores 1. Several times a day scores 5. Frequency matters more than people expect: a five-minute task done 30 times a week is worth more than a two-hour task done once a month.

2. How much time does it take, in total? Multiply the time per run by how often it runs, and be honest. Include the context-switching: finding the right spreadsheet, logging into the right tool, remembering where you left off. Under an hour a month scores 1. A full day a week or more scores 5.

3. How predictable is it? This is the one people skip, and it is the one that decides whether automation will actually work. If the task follows the same steps every time, with the same inputs, it scores 5. If every case is "it depends" and needs someone's judgement, it scores 1. AI tools have pushed this boundary a lot (they can now read messy emails or sort unstructured documents reliably), but a predictable process is still cheaper to automate and easier to trust.

4. What does a mistake cost? Here a higher score means more pain when things go wrong by hand. A typo in an internal note scores 1. A wrong invoice amount, a missed client follow-up or a stock count that leads to overselling scores 5. Automation is at its best when it removes the human errors that are expensive.

Step 3: Add it up and look at the top three

Add the four scores for each task. Anything above 15 is a strong candidate. Anything under 10 can probably wait.

Here is what that looks like for a typical small marketing agency:

The weekly reports win clearly. They happen often, they eat a whole afternoon, the steps are identical every week, and a wrong number in front of a client is embarrassing. Proposal drafts feel like the bigger pain, but they need judgement and change every time, so they are a worse first project.

Step 4: Sanity-check the winner

Before committing, ask three more questions about your top task:

• Do the tools involved have a way in? Most modern software (Google Workspace, HubSpot, Shopify, Xero, Notion and so on) can be connected. Older desktop software or tools that only export PDFs are harder, though rarely impossible.
• Does someone own the process? Automation works best when one person can say "yes, this is how it should work." If three people do it three different ways, agree on one way first. Sometimes that conversation alone saves more time than the automation.
• Can you measure the result? Write down how long the task takes today. In a month you will want to know whether it worked, and "it feels faster" is not a great answer.

What a good first automation looks like

The best first projects are rarely impressive. They are boring, frequent and clearly defined, and that is exactly why they work. They get finished quickly, they start saving time from the first week, and they show the team that automation is something that helps them rather than something that happens to them.

Once that first one is running, the second is easier to pick. You already know which tools connect, who owns which process, and what "done" looks like.

Want a second opinion?

If you've done the exercise and you're not sure your top task is the right one, or you'd like someone to estimate what it would take to automate, book a free 45-minute call. We'll go through your list with you and tell you honestly which items are worth automating and which aren't.