Should I hire an ops person or automate?

The arithmetic, honestly

Vendors present this comparison dishonestly, so here it is with the missing parts included.

The rows vendors leave out are the last three. An automation does not notice that a customer seems unhappy, does not suggest a better way of doing things, and does not cover for a colleague. Those are not edge cases; for many operations roles they are a large part of the value.

When hiring is right

• The work needs judgement. Deciding whether to make an exception for a good customer, spotting that a supplier's numbers look wrong, knowing this client needs a phone call rather than an email.
• Volume is unpredictable and lumpy. Automation shines on steady, repeated work. If the workload arrives in unpredictable bursts of varied shapes, a person absorbs that far better.
• Processes are still changing monthly. If the business is still working out how it wants to operate, you will automate a shape you are about to abandon. Hire, let the process settle, then automate it.
• The role covers a lot of ground. If someone would do twelve different things and automation addresses three of them, you still need the person.
• Relationships are part of the job. Supplier negotiation, client management, keeping a warehouse team informed.
• Nobody internal can own automation. A hire who is comfortable with tools like Make can be the better route, doing the work now and automating parts of it over time.

When automating is right

• High volume, rule-driven, stable. The same steps, many times a week, in the same order. This is the ideal case and it is worth more than any other factor.
• Errors are expensive. Data entry mistakes that cost money or reputation. A machine that is right every time beats a person who is right almost every time, on exactly this kind of task.
• The work is genuinely dull. This has a retention cost people underweight. Hiring someone to re-type invoices is hiring someone who will leave within a year, and then you pay recruitment again.
• It needs to happen outside working hours. Overnight order processing, weekend lead response, month-end runs at 2am.
• You cannot hire fast enough. Automation deploys in weeks against months for recruitment, and does not need to be found in a tight labour market.
• You would be hiring purely for capacity. If the honest job description is "do this repetitive thing more," that is a build, not a role.

Why it is usually both

The framing of "hire or automate" implies they compete for the same budget. In practice they are complements, and the businesses that get the most out of either do them together.

Consider a company drowning in order administration, weighing an operations coordinator at $65,000.

Hire alone: the new coordinator spends 60% of their week on data entry between the order system and the accounting package. They are competent and it gets done. Within eighteen months they are bored and looking, and the role has become a treadmill.

Automate alone: a $14,000 build removes the data entry. It handles about 85% of orders cleanly. The remaining 15% — the exceptions, the phone orders, the disputes — still need a person, and now nobody has been hired to be that person, so it lands back on whoever is nearest.

Both: the build removes the mechanical work; the coordinator handles exceptions, supplier relationships and improving the process. You get a role someone actually wants, and the automation has an owner who understands it. The build cost is roughly two and a half months of the salary you were going to spend anyway.

Which one first

If you are going to do both, sequence matters and the intuitive order is wrong.

Hire first if the process is undocumented. You cannot automate what nobody can describe. A good operations hire spends their first two months learning and writing down how things actually work — which is precisely the input an automation project needs. Automating first means paying someone to guess.

Automate first if the process is well understood and the hire would mostly be capacity. Here you already know the rules, and the build reduces the size of the role you eventually recruit for. You may find you need a part-time person rather than a full-time one.

One caution either way: do not hire someone and tell them their job is safe, then automate 60% of it six months later without discussing it. Whichever order you choose, be straight with people about the plan. We will tell you plainly which tasks disappear in any build we scope, precisely so you can have that conversation before it becomes a surprise.