It is a slightly awkward article for a services company to write, because the honest answer is often "hire someone." But hiring a firm when you should have hired a person goes badly for everyone, so it is worth being direct about the thresholds.
Separate execution from judgment
Two different needs get labeled "IT." Execution is the recurring work: tickets, devices, accounts, patching, vendor chasing. Judgment is the decisions: what to invest in, what to retire, what a contract should say, what risk is acceptable.
Execution scales with headcount and sites. Judgment scales with the size of the decisions in front of you. A company can need a great deal of one and almost none of the other.
Rough thresholds for execution
Ticket volume is the most reliable signal, if you have it. If you don't, headcount is a decent proxy.
- Under about 40 employees, single site: a part-time arrangement or an outsourced desk is almost always cheaper and more resilient than a hire.
- 40 to 120 employees: one internal generalist can work, but they will be a single point of failure. Vacation, illness and resignation are real operational risks.
- 120 to 300 employees: internal capacity starts to make sense, usually as a hybrid — your own people for on-site and relationship work, an outside firm for the harder problems and after-hours.
- Above 300, or above roughly eight sites: internal team, with outside help for specialties you cannot justify hiring for.
The single-point-of-failure problem
The most common failure we see is not a bad hire. It is a good one. A capable internal IT person accumulates knowledge that never gets written down, becomes the only person who can perform several critical tasks, and then leaves.
If you hire internally, budget documentation time explicitly and verify it exists. "Write the runbook" without a review date does not happen.
Judgment is a different purchase
A full-time CIO at this size is usually over-buying. But going without means large decisions — platform migrations, acquisitions, contract renewals, a list of fixes your insurer is demanding — get made by whoever is nearest, or not made at all.
A few days a month of senior involvement is generally the right shape until the roadmap is genuinely full-time work.
A test that usually resolves it
Ask two questions. First: if the person who currently handles IT resigned tomorrow, how long until it hurt? If the answer is measured in days, you have a continuity problem regardless of which route you choose.
Second: in the last six months, how many technology decisions above $50,000 did you make, and who advised on them? If the answer is "several" and "nobody," the gap is judgment, not headcount — and hiring a systems administrator will not close it.