Which Processes Should You Automate First? A Simple Way to Decide

This is a way to make that first pick deliberately. No tools to buy, no jargon: just two questions you ask about each candidate process, and a simple map that tells you where to begin.

The Mistake Almost Everyone Makes First

When people decide to automate something, they tend to reach for one of three things: the process that’s most visible (the one customers see), the one that’s most annoying in the moment, or the one that’s most technically impressive to automate. All three are traps.

The visible process is tempting because a win there feels like it matters most. But customer-facing work is usually full of judgment and exceptions, exactly the stuff automation handles worst, so you end up with a brittle system that breaks the first time reality doesn’t match the script. The annoying process feels urgent, but “annoying” and “costly” aren’t the same thing; the task that irritates you for two minutes a week is not where your hours are leaking. And the impressive process is the worst trap of all: complexity is fun to build and miserable to maintain, and a hard first project is the one most likely to stall before it ever pays off.

The right first automation is rarely the obvious one. It’s usually something boring, invisible, and quietly enormous: a process so routine nobody thinks to mention it, that happens to eat hours every single week. To find it, you stop going on instinct and score your options on two things.

The Two Questions That Decide Everything

For any process you’re considering, ask:

1. How much is it draining you today? Call this Drain.
2. How ready is it to be automated? Call this Suitability.

That’s the whole method. A great first automation scores high on both: it’s costing you real time and it’s genuinely easy to hand off to software. Most processes are strong on one and weak on the other, and knowing which is which is what keeps you from starting in the wrong place. Let’s make each question concrete.

Scoring Drain: How Much Is It Really Costing You?

Drain isn’t about how much a task annoys you; it’s about how much it quietly removes from your week. Three signals tell you, and you don’t need exact numbers, just honest estimates:

• Frequency: how often does it run? Many times a day is high drain. Once a quarter is low, almost no matter how tedious it is.
• Time per run: how long does one pass take, start to finish, including the waiting and the chasing? Five minutes is one thing; forty-five is another.
• Cost of getting it wrong: what happens when this process slips? A mistyped invoice, a missed follow-up, a double-booked appointment. Some processes are cheap to run but expensive to fumble, and that risk counts as drain too.

Multiply those three together in your head (frequency, time, and the cost of errors) and you get a rough sense of the weight a process is putting on your week. A short task done fifty times a day with real consequences when it slips is a heavy process, even though no single instance feels like much. That hidden multiplication is exactly what instinct misses.

Scoring Suitability: How Ready Is It to Automate?

A heavy process is only worth starting with if software can actually take it over cleanly. Suitability measures that, and again there are three signals:

• Is it rules-based? Can you describe how to do it as a clear set of “if this, then that” steps? If the right action is always determined by the inputs, software thrives. If it depends on reading the room or making a call, it doesn’t.
• Is it the same every time? A process that runs identically on every pass is easy to automate. One riddled with special cases and “well, except when…” is not, at least not yet.
• Are the inputs already digital? Does the information arrive in a form a computer can read (a form, a spreadsheet, an email, an app), or does it live on paper, in someone’s head, or in a verbal handoff? Digital, structured inputs make automation straightforward; messy or offline inputs add a whole project before you even begin.

A process that’s rules-based, consistent, and fed by clean digital inputs is highly suitable: software can take it almost as-is. A process that depends on judgment, varies constantly, and relies on information trapped on paper is low-suitability, even if you’d love to be rid of it.

The Map: Plot Drain Against Suitability

Now put the two scores together. Every process you’re weighing lands in one of four quadrants:

	Low Suitability	High Suitability
High Drain	Redesign or AI-assist	⭐ Automate first
Low Drain	Leave it alone	Quick wins (later)

• High Drain + High Suitability → start here. Heavy enough to matter, clean enough to hand off. This is your first automation. It pays off fast and proves the idea to everyone watching.
• High Drain + Low Suitability → redesign, don’t automate yet. These hurt, but they’re not ready, usually because they’re full of judgment or fed by messy inputs. The move isn’t to force automation; it’s to simplify or restructure the process first, or use AI to assist the human rather than replace them. Trying to fully automate these on day one is how first projects die.
• Low Drain + High Suitability → easy, but later. Genuinely simple to automate, but the payoff is small. Tempting because they’re quick (and a quick win can be worth it for morale), but don’t let them jump the queue ahead of the top-right.
• Low Drain + Low Suitability → leave it alone. Not worth the effort, not ready anyway. Ignore it and feel no guilt.

The discipline is simple: always start in the top-right corner. Everything else is a distraction dressed up as an opportunity.

What This Looks Like in Practice

Make it concrete. Imagine a mid-size accounting firm (fictional, just to show the method) weighing three processes for its first automation. (No real numbers here; the point is the reasoning, not the figures.)

Process A: answering client questions by email. High drain: it eats hours every day. But low suitability: every question is different, the right answer depends on the client’s specific situation, and a wrong answer is costly. This is judgment-heavy. It lands in redesign / AI-assist: maybe a tool that drafts a first reply for a human to check, but not full automation.

Process B: formatting the office’s monthly newsletter. Highly suitable: it’s the same template every time, all digital. But low drain: it happens once a month and takes a single afternoon. A quick win at best. Not where you start.

Process C: pulling receipts from incoming emails, entering them into the books, and flagging anything missing. It runs constantly, takes real time, and an error means a client’s books are wrong: high drain. And it’s rules-based, identical every time, with digital inputs: high suitability. Top-right corner. This is the firm’s first automation, and it isn’t the one that first came to mind. The flashy one (A) was the wrong place to start; the boring one (C) was the gold.

That’s the pattern almost every time: the best first automation is the unglamorous process hiding in plain sight.

Three Traps to Avoid Even After You’ve Chosen Well

The matrix points you at the right process. These three things still sink good first projects:

1. Automating a broken process. If a process is a tangled mess by hand, automating it just makes the mess run faster and more reliably. Fix or simplify the process first, then automate the clean version. Automation is an amplifier: it amplifies good processes and bad ones equally.
2. Starting with the hardest thing to prove you’re serious. Ambition is not a strategy. A hard first project maximizes the chance of stalling before anyone sees a result. Your first automation’s real job is to build belief, so bias toward the clear win you can finish, not the moonshot.
3. Forgetting the people who run it. The person doing a process by hand today knows every exception and edge case you’ve never written down. Leave them out and your automation will miss them all. Bring them in and they become the ones who own it, handle what software can’t, and find the next thing worth automating. Automation done well gives your people better work, not less of it.

Frequently Asked Questions

How many processes should I try to automate at once?

One. Especially the first time. The goal of your first automation isn’t scale: it’s a clear, finished win that proves the idea and teaches you how this works in your specific business. Nail one process in the top-right corner, feel the time come back, and the next candidates become obvious. Trying to automate five things at once is the fastest way to finish none of them.

What if my most painful process isn’t very automatable?

That’s common, and it’s exactly what the matrix is for. A high-drain, low-suitability process (usually one full of judgment or fed by messy inputs) isn’t your first automation: it’s your first redesign. Simplify the process, get its inputs into digital form, or use AI to assist the human handling it rather than replace them. Once it’s cleaner, it may move into the top-right. Forcing it before then is how first projects fail.

Do I need to measure exact hours and costs to use this?

No. Honest estimates are enough to rank your options, and ranking is all you need to pick a starting point. Is this process daily or monthly? Minutes or hours? Cheap or costly when it slips? You can answer those well enough from gut and a week of paying attention. Precision matters later, when you’re measuring the result, not now, when you’re just deciding where to begin.

We have no technical team. Can we still do this?

Yes. The deciding part of this is a business judgment, not a technical one. You know your processes better than any engineer does, and this method runs entirely on that knowledge. The scoring tells you what to automate; the how is where a partner comes in. Walking in already knowing your top-right process makes that conversation far shorter and far cheaper.

Start in the Top-Right Corner

The reason most automation efforts stall isn’t the technology: it’s starting in the wrong place. Pick by instinct and you’ll grab the visible, the annoying, or the impressive, and most of those are traps. Pick by Drain and Suitability and you’ll find the boring, invisible, quietly enormous process that was the right answer all along. Score your options, plot them on the map, and begin where the weight and the readiness overlap. That first deliberate win is what turns “we should automate something someday” into a business that actually does.

Want help finding the top-right process in your business, and turning it into a win? Talk to our team → and we’ll help you map it.