A sensible first automation: how to spot the right task to start with

Your first automation should be boring, frequent and low-risk. Here is how to spot the right task to start with, and which ones to leave alone for now.

The most common automation mistake is starting with the wrong task. People tend to pick the job that annoys them most, or the one that sounds most impressive. Neither is a good filter. The best first automation is usually something small, frequent and forgettable.

The four traits of a good first task

A task is a strong candidate when it ticks most of these boxes.

• It happens often. Daily or several times a week. Frequency is what turns a small saving into a real one.
• It follows the same steps each time. Rules you can write down beat judgement calls you can only feel.
• It is low-risk if it hiccups. A mislabelled folder is annoying. A wrong invoice is not. Start where mistakes are cheap.
• It has a clear start and finish. “When a form is submitted, do X, then Y” is automatable. “Handle the client relationship” is not.

Notice that “how much I hate it” is not on the list. Some hated tasks are hated precisely because they are messy and full of exceptions, which makes them hard to automate well.

Good starting examples

These come up again and again in small businesses and tend to work.

• Copying enquiry details from a web form into your spreadsheet or CRM.
• Sending a standard acknowledgement email when someone gets in touch.
• Saving email attachments into the right folder automatically.
• Creating a draft invoice from a completed job record.
• Reminding you when a task has sat untouched for too long.

None of these are glamorous. That is the point. Boring automations are reliable automations.

Tasks to avoid at first

Leave the hard ones until you have a couple of wins behind you. Avoid anything that needs a human to make a genuine decision partway through, anything where the steps change depending on mood or context, and anything that touches money or compliance until you are confident. Those can absolutely be automated later, but they are poor first projects because the cost of getting it wrong is high and the setup is fiddly.

Work out what a saving is really worth

Before you build anything, do quick sums. If a task takes ten minutes and happens twenty times a week, that is over three hours a week, or the best part of a working month across a year. That is worth an afternoon of setup. If it takes ten minutes but happens twice a month, it almost certainly isn’t. This simple test kills a lot of tempting but pointless projects.

If you find it hard to see where your time actually goes, an automation audit is a structured way to spot the frequent, low-value tasks that are quietly eating the week. Once you have a candidate, the build itself is usually straightforward process automation: a trigger, a few steps, and a sensible fallback if something looks odd.

Start small and prove it

Pick one task. Automate it. Then watch it for a week and check it actually behaves. A single working automation you trust is worth more than five half-finished ones you have to babysit. Once one is running quietly in the background and you have stopped thinking about it, you will know you picked the right task, and you will have a much better instinct for choosing the next.

The goal is not a clever system, it is one boring job you never have to do again.