How to Decide What to Automate First

For a lot of growing businesses, automation starts as a feeling before it becomes a plan.

Something in the business feels repetitive. Admin is building up. People are copying information from one place to another. Updates are being chased manually. Work slows down because too many small tasks depend on someone remembering, checking or moving something forward.

At that point, the natural reaction is to say:

“We need automation.”

And often, that is true.

The problem is that many businesses then jump straight from that feeling to a tool, without making a clear decision about what should actually be automated first.

That is where time and money get wasted.

Automation can be incredibly useful, but only when it is applied in the right place, for the right reason, and in the right order.

Automation is most effective when it removes repeatable drag

The best first automation opportunities are usually not the flashiest ones.

They are the tasks that happen frequently, follow a recognisable pattern, and create drag because they still rely on manual effort.

That might include:

• moving information between systems
• sending routine notifications
• routing forms or enquiries
• assigning follow-up tasks
• updating statuses
• triggering reminders
• creating standard internal handovers

These tasks often do not look strategic on the surface.

But they matter because they happen again and again.

And when something repetitive sits inside a busy workflow, even a small inefficiency can become expensive over time.

That is why the best first automation is often the one that removes a lot of repeated friction, not the one that sounds most impressive in a meeting.

The wrong starting point is “what can this tool do?”

This is one of the most common mistakes.

A business buys or trials an automation platform, then starts looking for ways to use it.

That reverses the logic.

The question should not be:

“What can this tool automate?”

It should be:

“Where is the business repeatedly losing time on work that follows a stable pattern?”

That difference matters.

Starting with the tool often leads to automating what is easiest to automate, rather than what is most valuable to automate.

Starting with the workflow leads to much better decisions.

What makes something a good first automation candidate?

A strong first automation candidate usually has most of these characteristics.

1. It happens often

Frequency matters.

If something only happens once in a while, it may not be worth addressing first.

But if it happens daily or several times a week, even a modest saving can add up quickly.

This is one reason why repetitive admin and internal handovers are often strong starting points. They may feel small, but they occur constantly.

2. It follows a clear pattern

Automation works best when the task is stable enough to define.

If the steps are always changing, if the conditions are unclear, or if no one agrees on how the process is meant to work, automation is usually premature.

A good first automation opportunity is one where you can clearly say:

• what triggers it
• what should happen next
• what the rule is
• what the output should be

Clarity makes automation more reliable.

3. It involves low judgement

Not everything in a business should be automated.

Some tasks depend on experience, nuance, context or commercial judgement. Those are usually not the best place to begin.

The best first automation candidates are tasks where the decision-making is limited and the process is mostly procedural.

For example:

• send this update when that stage changes
• route this enquiry to the right person
• create a reminder if no reply has been received
• copy this information from one system to another
• create a task when a form is submitted

These are the kinds of tasks automation handles well.

4. It creates visible friction today

A lot of businesses could automate many things, but not all of them are worth doing first.

The strongest opportunities are the ones already causing pain:

• delays
• admin burden
• missed follow-up
• repeated checking
• inconsistent handovers
• tasks people complain about regularly

If the team is already feeling the drag, and the task is stable and repetitive, that is usually a strong sign you are looking in the right place.

5. The risk of getting it wrong is manageable

Not every workflow should be your first experiment.

If a task is highly sensitive, customer-critical, legally risky, or difficult to undo when something goes wrong, it may not be the smartest place to begin.

A better first automation is one where:

• the process is easy to test
• the outcome is easy to verify
• the risk is contained
• the business can learn safely

That allows you to build confidence before moving on to more complex automations.

Good examples of what to automate first

Every business is different, but some common first automation candidates include:

• routing new enquiries to the right person or team
• creating follow-up reminders after a form submission or sales action
• sending internal notifications when a job changes stage
• copying structured information between systems
• creating standard tasks when a project or order starts
• prompting missing information before work moves to the next stage
• sending routine customer updates triggered by clear status changes

These are not glamorous.

But they are often exactly the kinds of things that reduce drag quickly.

What not to automate first

It is just as important to know what not to start with.

The wrong first automation usually has one or more of these characteristics.

1. The process is already messy or unclear

If the workflow itself is confused, automation will not fix that.

It may actually make it worse by hardwiring confusion into the system.

Before automating, the business needs enough clarity to understand:

• what the process is
• who owns it
• what should happen
• where the exceptions sit

If that clarity is missing, fix the workflow first.

2. The task depends heavily on judgement

Tasks involving commercial nuance, context-heavy decisions, negotiation, or inconsistent inputs are rarely the best first place to automate.

That does not mean they should never be supported by automation or AI.

It just means they are not usually the lowest-risk, highest-confidence place to start.

3. The volume is low

If a task happens rarely, the return may be too small to justify putting it first.

That does not make it a bad automation forever. It just means there are probably stronger opportunities elsewhere.

4. The inputs are unreliable

If the process depends on inconsistent, incomplete or messy information, automation can become fragile very quickly.

Bad inputs create bad outputs.

A lot of failed automation projects are really data-quality or process-discipline problems in disguise.

5. It sounds exciting, but does not remove much drag

This is more common than people think.

Sometimes businesses pick an automation project because it sounds innovative, not because it removes a meaningful operational burden.

That usually leads to activity without much real gain.

Your first automation should create a useful improvement people can actually feel.

A simple way to prioritise automation opportunities

If you are deciding what to automate first, it helps to score each candidate against a few practical criteria:

• Frequency — how often does it happen?
• Time cost — how much effort does it currently take?
• Friction — how annoying or disruptive is it?
• Clarity — how well-defined is the process?
• Risk — how safe is it to automate?
• Impact — if improved, how much better would the workflow feel?

You do not need a perfect spreadsheet model.

Even a simple discussion around those criteria will usually help reveal which automation opportunities are genuinely worth prioritising.

Why many businesses struggle to choose

A lot of businesses do not struggle because there are no automation opportunities.

They struggle because everything feels inefficient at once.

When that happens, it becomes difficult to tell the difference between:

• what is actually causing the drag
• what is merely visible
• what is easy to automate
• what is important to automate
• what should be fixed through process, not technology

That is exactly why choosing what to automate first can feel harder than it sounds.

The challenge is not usually lack of options.

It is lack of clarity.

The best first automation is often not the biggest one

This is worth stressing.

Your first automation should not necessarily be the most ambitious project in the business.

It should usually be the one that:

• solves a real pain point
• is stable enough to automate properly
• improves the day-to-day operation
• creates confidence in the process
• teaches the business what good implementation looks like

That is how automation becomes useful and repeatable, rather than another abandoned initiative.

Why this matters before investing further

If a business gets its first automation decisions wrong, two things usually happen.

First, the automation underdelivers.

Second, confidence drops.

People start to think automation is overhyped, awkward, or more trouble than it is worth.

In reality, the issue is often not automation itself.

It is poor prioritisation.

Getting the first few decisions right matters because it shapes whether automation becomes a trusted capability or just another disappointing experiment.

The best starting point is understanding where the drag really is

This is why many growing businesses benefit from diagnosing the workflow before deciding what to automate.

If the business is not yet clear on:

• where time is being lost
• which manual steps are truly avoidable
• which handovers are weak
• what is causing the friction
• what the process should ideally look like

then it is very easy to automate the wrong thing first.

That is why a Business Efficiency Audit is often the right starting point.

It helps identify:

• where the manual drag sits
• which processes are stable enough to improve
• where automation can actually help
• what should be fixed first

That makes automation decisions more commercial, more practical and far more likely to work.

Final thought

Automation is powerful, but only when it is pointed at the right problem.

The best first automation is rarely the most exciting idea. It is usually the task that is frequent, clear, low-judgement, and already causing visible friction.

If your business knows it needs automation but is not yet sure where to begin, do not start with the tool.

Start with the workflow.

That is usually where the real answer is.

Not sure what to automate first?

If your team is spending too much time on repetitive admin, manual handovers or avoidable coordination, but you are not yet clear which automation opportunity matters most, start with a Business Efficiency Audit.

It helps uncover where the drag really sits, what is worth improving first, and where automation can create the biggest practical gain.

Explore the Business Efficiency Audit