How do I know whether I need to hire?

Run the test before you write and post the ad. Hiring is the right answer often enough that it's worth doing well, and the wrong answer often enough that it's worth checking first. The check takes about an hour and could be spread over a fortnight.

The question underneath "do I need to hire" is usually "why is there more work than there are hours". There are four honest answers: the work is genuinely growing, the work is being done twice, the work doesn't need doing at all, or the work routes through you because nobody else has been told how to do it. Only the first one is a hiring problem.

Key takeaways

• Hiring someone into a process that doesn't exist gives you a second person doing unstructured work from memory, with no way to check it.
• Count the hours before you count the salary. Two weeks of a simple log could tell you what the role would actually be working on.
• Every recurring task has three possible homes: a system, a tool, or a person. Test them in that order.
• A part-time hire into a written system usually beats a full-time hire into a system that only lives in your head.
• If cash is thin or demand is lumpy, the answer may be not yet.

Step one: find out what the work actually is

For two weeks, log the work that is spilling over. You don't need fancy software for this. A note on your phone with the task, roughly how long it took, and whether it had to be you is enough.

At the end of the fortnight, sort the log into four piles.

• Repeating work with a right answer. Quoting, onboarding, invoicing, ordering, scheduling. Same shape every time.
• Repeating work with judgement in it. Pricing an unusual job, handling an unhappy customer, deciding what happens when two jobs clash.
• One-off work. The website, the new supplier, the insurance renewal.
• Work that shouldn't exist. Reports nobody reads, double handling between two systems, a step that survives because of a problem you fixed two years ago.

A lot of business owners are surprised by the size of the fourth pile. You can start there by getting rid of it.

Step two: run the three tests, in order

Take the largest remaining pile and ask three questions in this order. The order matters, because answering them out of order is how a business can end up paying a salary to maintain a mess.

1. Can it be written down? If the task has a right answer, the first move is a written standard: what good looks like, the steps, and who decides when something is unusual. This is cheap, and it's what makes the other two options work.

2. Can a tool handle part of it? Once it's written down, some of it will be obviously mechanical. Reminders, scheduling, moving data between two systems, a first draft. This is the point at which a tool helps, because you now know what the tool is meant to produce.

3. Does the rest need a person? What survives both passes is the actual role. It's usually smaller than the job you were about to advertise before doing this exercise, and much clearer to describe.

There's rigorous evidence behind putting the system first. A randomised controlled trial of owner-managed firms, published in the Quarterly Journal of Economics, found that when better systems and information flows were put in place, owners began delegating decisions they had previously kept to themselves, and that the reluctance to hand work beyond the owner had itself limited how large those firms could grow. The system came first and the letting go followed.

Step three: check the numbers honestly

Three figures decide whether you can hire, separately from whether you should.

• The full cost. Salary plus superannuation, insurance, equipment, software seats, and payroll tax if you cross the threshold. Then add your own time. A first hire takes real hours to train and mentor, usually across the first eight to twelve weeks.
• The work that pays for them. Not the work you hope will arrive. The work already contracted, or a pipeline you would still believe in if the next two months went quiet.
• The buffer. How many months of that salary you could cover if a large customer left. In the 2025 COSBOA and Commonwealth Bank study of more than 800 Australian small business owners, 60% couldn't pay themselves consistently and around a quarter regularly dipped into personal savings. Thin reserves are common, and they change what a hire risks.

If all three hold, it could be a good time to hire. If the third doesn't, the answer is more likely to be not yet, and the systemising you did in step two has bought you time to get there.

When the answer is no, or not yet

Some honest cases where hiring is the wrong next move.

• The work is seasonal. A busy quarter isn't when you want to hire. It's best to contract, subcontract, or extend hours through the peak.
• You can't describe the job. If you can't write a page explaining what good looks like, a new person won't be able to guess it either.
• The bottleneck is a decision. If everything waits on your approval, a second person joins the same queue. See How do I delegate without losing control?.
• Cash is tight this quarter. A hire you have to reverse costs more than the work it absorbed.

What to do first if the answer is yes

Write the system before you write the ad. One page per core task: what good looks like, the steps, the limits inside which the person decides for themselves, and how you will check. Then hire into that page. The first month of a new person's time goes to running the system and improving it, which also tells you quickly whether they can.

For the fuller version of the ordering question, What should I systemise first? works through which system to build before the others. If the pressure came on suddenly, What do I do when the business grows faster than I can keep up? covers the same ground under load.

And if you want support to build your system, 25eight's Next-Level Growth Program would be a great fit.

Frequently asked questions

Should I hire a generalist or a specialist first? This is highly dependent on your business and needs but we generally suggest that a generalist is best, if what you have is volume across several repeatable tasks. A specialist may be best, if what you have is one job you can't do well yourself. The two-week log answers it. Work spread across five kinds of task is generalist work.

Is a part-time hire a compromise? For most owner-led businesses at a particular size it's the sensible first step. Part-time into a written system gives you a real test of whether the role works, at a cost you can reverse if the pipeline changes.

Should I use AI instead of hiring? Sometimes, for the mechanical parts. Adoption in Australia is lower than the headlines suggest. The Australian Bureau of Statistics found about 12% of Australian businesses used AI in 2024 to 2025, and 11% of small businesses. The more useful figure in the same release is that innovation-active small businesses adopted at 19%, against 4% for those that weren't. Tools follow capability rather than substituting for it. Automate the written steps and keep the judgement with a person. We can help you determine if you have opportunities in our AI Opportunity Mapping Session

What if I have hired before and it didn't work? Look at whether a written standard for the role existed at the time. A hire without one is being asked to reproduce your judgement by observation, which almost nobody can do. That's a design problem in the role rather than a character problem in the person.

How do I know the pipeline is real? Count only work that is contracted or has a date. Then ask what happens if your largest customer leaves this quarter. If the answer still supports the salary, it's real enough.