Before you add a salary, ask whether the task needs judgement or just repetition. A simple test sorts the work into what deserves a person and what deserves a system, and tells you which to fund first.
Hire or automate first: the short answer
Hire first when the work needs judgement, relationships or accountability that a customer can feel. Automate first when the work is repetitive, rule-based and already done the same way every time. Most small businesses have both kinds of work tangled together in one overloaded role, so the best first move is usually to automate the repeatable part and hire for what remains.
That order matters because a new salary is a fixed monthly commitment, while a system is a one-time build with a small running cost. Getting the sequence wrong is expensive in either direction: you pay a person to retype data, or you automate something that needed a human touch.
The real price of the RM4,000 decision
A monthly salary of RM4,000 is a useful round number, not a quote, but the real cost of a hire is higher than the figure on the offer letter. Employer EPF contributions are 13 percent of monthly wages up to RM5,000 and 12 percent above that, before SOCSO and EIS, leave, equipment, a desk, training and the manager's time all sit on top.
A hire also takes months to become productive and stays a cost whether the work is there or not. A system works the other way: more build effort up front, almost nothing per task afterwards, and no ramp-up. Neither is cheaper in general. What decides it is the shape of the work.
Three kinds of work, three different answers
Sort every recurring task into one of three groups. Judgement work involves exceptions, negotiation, taste or trust: handling an upset customer, pricing an unusual job, deciding which lead is worth a call. It deserves a person. Routine work follows rules: copying enquiry details into a quote, sending reminders, updating a spreadsheet, filing documents. It deserves a system. Mixed work has a routine skeleton with a judgement step in the middle, such as qualifying a lead, where software can gather and score the information and a person makes the call.
Most of what is called admin falls in the second and third groups. That is why the person you were about to hire for admin often turns out to be a person you would have asked to spend most of the day copying and pasting.
How to decide whether to hire or automate: a one-week test
Do not decide from a feeling of being overloaded. Run a one-week log. Everyone on the team notes each recurring task, how long it took, how often it happens and whether it needed a decision. At the end of the week, you have a list you can sort.
For each task, ask four questions. Does it follow the same steps most of the time? Could a new colleague do it from a written checklist? What is the cost of a mistake, and who notices? Would the customer care that a person did it? Tasks that get yes, yes, low and no are candidates for automation now. Tasks that get no, no, high and yes need a person. Anything in between is mixed work, and the answer is a system that prepares and a person who decides.
Then add up the hours. If the automatable tasks come to a large share of a full-time role, build the system first and see how much room it frees. If the remaining judgement work still fills a role, hire, and hire for that. The new person then starts with a job worth doing instead of an inbox to clear.
Five rules for the hire-or-automate decision
- Automate the task, not the person. Write down the process as it runs today before touching software. If two people do it differently, agree on one way first; automating an inconsistent process only produces inconsistent results faster.
- Hire for judgement and relationships, then protect their time. A good salesperson or account manager should not spend mornings entering data. Remove the routine work around them and the same salary produces more of what only they can do.
- Count the cost of delay, not just the cost of labour. A lead answered late, an invoice sent late, a follow-up forgotten: these are losses that a salary does not fix by existing. Check where work waits, because systems remove waiting while hiring often adds a handover.
- Start with one process and measure it. Pick the task with the clearest rules and the highest weekly hours, automate it, and compare hours before and after. If the saving is real, move to the next; if not, you spent a small amount to learn something.
- Keep a human exit on every automated flow. Customers and staff need a clear way to reach a person when the system meets something it was not built for. A system that cannot hand over gracefully will cost you more in trust than it saved in time.
Hire or automate: common questions
- Is it cheaper to automate than to hire? — It depends on the work. For repetitive, rule-based tasks, a one-time build with a small running cost can undercut an ongoing salary over time. For work that needs judgement or relationships, automation does not replace the person, so the comparison does not apply.
- What tasks should I never automate? — Anything where the customer is paying for your judgement or your relationship with them: sensitive complaints, negotiation, pricing exceptions and decisions with real consequences. Automation can prepare the information for these moments, but a person should make the call.
- Should a ten-person company hire an admin assistant or build automation? — Log a week of the admin work first. If most of it is copying data, sending reminders and chasing documents, automate that and hire later for whatever remains. If it is mostly coordination, handling exceptions and talking to people, an assistant is the right hire.
- Can I automate first and still hire later? — Yes, and it usually makes the later hire better. The new person inherits clean processes, clear data and a defined role, and spends their time on the work that actually needs them.
Keep reading
Build something
worth trusting.
Tell us what's slowing your business down. We'll show you the system that fixes it — and how fast.