There is a pattern we meet in almost every small business we audit, and it goes like this: the Wi-Fi, the backups, and the passwords all live in one person's head. Usually the owner's head, or the head of the one employee who "knows the systems." Nobody wrote any of it down, because writing it down felt like a low priority on a Tuesday, and now the business depends on that person remembering everything.
This is not a technology problem yet. It is a memory problem with technology attached. Here is what it looks like when it goes wrong, and what the fix actually costs.
The three failure modes
The person leaves. Whether they quit, get sick, or take a vacation at the worst possible moment, the knowledge goes with them. The new person's first week is archaeology: finding accounts, guessing passwords, discovering that the backup software was configured by someone who is no longer answering calls.
The person stays, and the knowledge quietly goes stale. The Wi-Fi password that has not changed since 2019, the vendor account that survived three name changes, the server that "has always been like that." It is not forgotten — it is just never questioned, and nobody can tell you why it is the way it is.
The person is fine, and the backups are not. This is the one that hurts. The backups were set up years ago, nobody has ever checked them, and the monitoring — if there is any — stopped sending alerts months back. The first sign of trouble is the moment you need a restore and there is nothing to restore.
Why it is always a surprise
Because nothing about this fails loudly. The office runs. The invoices go out. The shared drive works. The failure is in the second derivative: not that things break, but that you no longer know what would happen if they did. An IT stack in this state is like a smoke alarm with dead batteries — indistinguishable from a working one until it matters.
What the fix looks like
The fix is not a bigger IT budget. It is documentation and verification:
- An inventory. Every machine, account, vendor and subscription, written down, with an owner.
- Tested backups. Not "backups are running" — restores we have actually performed, on a schedule, with a log.
- A runbook. What to do when X happens, in language a non-specialist can follow at 2 a.m.
- A password home. One place that holds the keys, protected properly, so the knowledge stops being a person.
That is the core of what we do in IT infrastructure work, and it is why our infrastructure audit starts with the parts nobody wants to look at. The deliverable is not a report that gathers dust; it is the diagram, the runbook and the restored backup — the things your next hire, or your next contractor, can actually use.
The honest pitch
If your systems live in one person's head, the cheapest insurance in the business is to get them out of there. If the person is you, this applies double: you cannot be both the owner and the single point of failure. The audit is a fixed price and a complete deliverable — you keep everything whether or not you keep us. And if your stack is actually fine, we will say so, which is a report worth having on its own.
Start with a 30-minute call. Bring the list of things you hope never break.