When Is the Right Time to Automate? A Three-Question Test

Owners usually ask when they'll be "big enough" to automate. It's the wrong question. Plenty of $30M companies automate the wrong things, and plenty of $5M shops automate exactly the right one. The right time to automate isn't a revenue milestone. It's a property of the task in front of you.

Run any candidate through three questions.

Question 1: Is the task high-volume?

Automate tasks you do a lot. The return on automation comes from repetition, so the first filter is frequency: does this happen daily, or many times a week?

A task you do twice a year isn't worth automating even if you could. A task you do forty times a day is quietly funding a part-time job nobody hired for. Volume is what turns a small per-instance saving into real reclaimed capacity.

Question 2: Is the task rule-based?

Automate tasks where the steps are the same every time and there's no judgment call in the middle. If a task follows a predictable path — this input, that action, this result — it's a strong candidate.

If the task requires reading a situation and deciding, it isn't ready yet. You can often automate the routine 80% and route the genuine exceptions to a person. But if the whole task is judgment, leave it with the human.

Question 3: Does everyone already agree on how it's done?

This is the question people skip, and it's the one that sinks projects. Automate only what your team already does the same way. If people still argue about how a task should work, you're not ready.

Here's why it matters: automation doesn't fix a messy process. It amplifies whatever process it sits on. Hard-code a workflow nobody agrees on and you don't get clarity — you get the confusion running faster, with a tool's name on it.

What if I get a "no" on question three?

A "no" on agreement is not a reason to give up — it's the actual work, surfaced early. Before you automate, get the three people who do the task in a room and settle how it should run. Write it down. Now you have something worth automating.

Most stalled automation projects didn't fail at the technology. They failed because someone automated a process that was never agreed on in the first place.

So what should I automate first?

Automate the task that earns three clear yeses: high-volume, rule-based, and already agreed upon. That's free money sitting on the floor — the lowest-risk, highest-return automation you own.

Pick that one. Prove it runs without babysitting. Trust it. Then do the next one. You don't need an automation platform or a transformation initiative. You need one boring, repeatable task taken off a person's plate — and the discipline to settle the process before you speed it up.