Should I Hire Someone or Set Up Automation?

When the workload grows, one sentence shows up on your desk: "At this rate I will need to hire someone."

That same week you see an ad for automation and a second sentence appears: "Maybe software could handle it."

Both sentences can be reasonable. The trouble is that they look like answers to the same question. They are not. This article is about separating the two.

The question is usually framed wrong

"Hire or automate" assumes the two are interchangeable. In most small businesses they are not.

The better question is this: is the work piling up on your desk repetitive work, or work that requires judgement?

Repetitive work follows the same steps in the same order. Copying an incoming message into a spreadsheet, filing an invoice, sending an appointment reminder, answering the same five questions with the same five answers. People get tired doing this, lose focus, and skip steps.

Work that requires judgement is different every time. Hearing discomfort in a customer's voice, negotiating with a supplier, calming down a complaint, deciding on an exception. People are strong here and software stays weak.

In most businesses the pile is a mix of both. And a decision made before sorting the pile tends to come out wrong.

What each option actually costs

You need to run this comparison with your own figures, because it varies by sector and by city. Listing the line items without forgetting any of them can make the work easier.

Line items for hiring:

• Net salary
• Insurance and tax burden
• The recruitment process and the time it consumes
• The training period, meaning the weeks that pass before output begins
• Desk, computer, phone line, software licence
• Work pausing during leave and sick days
• The chance of the person leaving and the process starting over

Line items for automation:

• Setup fee, paid once
• Monthly maintenance and support
• Subscriptions for the services being used
• The time you need to set aside during setup
• Redevelopment when something needs to change
• Who looks after the system when it stops

When you put the lists side by side, one pattern usually appears: the cost of a hire repeats every month and grows over time. The cost of automation gathers mostly at the start, then drops and settles.

This does not mean automation is cheaper. It means the shape of the cost over time is different. In a business with tight cash flow, a cost concentrated at the start can be hard, while a fixed monthly figure may feel manageable. The reverse is also possible.

Which work suits which option

Try sorting your pile into three boxes.

Box one, work that suits automation. Tasks whose rule can be written down. If you can describe it with the sentence "when this arrives, do that", it is a rule. Appointment reminders, invoice filing, stock alerts, answers to frequently asked questions, writing a submitted form into a table. Automation does not get tired here and does not forget. Where a manual spreadsheet stops keeping up is covered in when Excel stops being enough for customer tracking.

Box two, work that suits a person. Tasks whose rule cannot be written down. Complaint handling, negotiation, building a new customer relationship, deciding on an exception, delivering bad news well. Building software here can be money spent for nothing.

Box three, work where both run together. This is usually the largest box and it is the one most often overlooked. An example: the system records the incoming request, classifies it, queues it by urgency, then hands it to a person. The person does not start from zero, they start with a prepared file. In this arrangement one person may keep up with more work than before.

Box three matters because the decision is often not "hire or automate" but "can I relieve the person I already have before hiring another one". A close topic is covered in what to do when work stops while you are away.

What automation does not cover

To be straight about it, automation does not fill every gap.

Building a new customer relationship is human work. Being on site is human work. Noticing a problem and solving it before anyone reports it is human work. If your business is squeezed on this side, software will not bring relief, only a tidier version of the same squeeze.

It is also worth counting the maintenance automation needs. A system that has been built does not carry itself forward. When the process changes it needs updating, when a service shuts down it needs reconnecting, and it needs checking now and then. If nobody agreed who would do this before setup, the system can quietly start working incorrectly a few months later. The same theme runs through why automation projects fail.

What hiring does not cover

Let us apply the same honesty to the other side.

When a new person arrives the workload does not drop immediately. First it rises, because teaching takes time and you are the one teaching. Output generally shows up a few months later. Timelines and expectations are measured in how long automation takes and when it starts to pay off.

Hiring does not repair a broken process, it only puts a person on top of it. This second point matters more than the first. If you are entering the same information in three places, the second person will also enter it in three places. The error count does not fall, it can double.

This is why the right order in many businesses is: simplify the process first, then add people if still needed.

Five questions before deciding

Answer these yourself. Writing the answers on paper tends to make things clearer.

1. Could someone else do this task if it were described to them? If yes, and if the description can be written down, automation is worth a look.

2. Is the volume steady or does it swing? Steady volume supports both a person and software. With swinging volume, a fixed salary can become a burden in the quiet months.

3. What does an error cost? If a missed appointment loses a customer, a system that does not forget has value here. If the cost of an error is low, so is the urgency.

4. Who does this work today and how much of their time does it take? Keep a rough log for a week. Deciding from memory can mislead you, because the task that eats the most time is usually not the most visible one.

5. Will this work grow in six months? If it will, solving it with a person today may bring the same question back in six months.

Three common mistakes

Trying to automate everything at once. Large builds take a long time, and patience runs out before long builds finish. Picking a single problem and finishing it usually goes more smoothly.

Moving to a system before writing the process down. The flow in your head and the flow that actually happens are often not the same. When a process that has not been written down gets automated, the errors get automated too.

Nobody owning the system after setup. If it is not clear from the start who checks the system and who to contact when it breaks, the system can fall out of use without anyone noticing.

Summary

The question is not "hire or automate". The question is: how much of your pile is repetition and how much is judgement?

If the repetition side is heavier, looking at automation can make sense. If the judgement side is heavier, looking for a person may be the better move. Most businesses have both, and the right order is to simplify the repetitive side first, then measure the need for a person again. If you are weighing a ready-made solution against something built for you, custom software or an off-the-shelf package may help.

When that order is followed, sometimes a second person turns out not to be needed. Sometimes they are still needed, but they arrive into a role with a clearer definition.