Every January a small business owner opens a spreadsheet they have not touched since last January, and discovers it is still load-bearing. The invoice tracker. The staff rota. The one tab that reconciles two systems that were supposed to talk to each other by now. It works, in the sense that nothing has fallen over. That is not the same as working.
January is the only month most small operations get a genuine look at themselves. The holiday rush is over, the year-end numbers are fresh, and nobody is mid-project. It is also the month the audit gets skipped, because the quiet is the first quiet in months and the temptation is to spend it on anything else.
What actually goes wrong
The failure mode is not that a system breaks. It is that a system silently becomes one person’s job. Someone knows the Tuesday export has to happen before the Wednesday invoice run, and that if a customer name has an ampersand in it the import fails and you fix it by hand. None of that is written down. It is not a process — it is a habit with a person attached.
You find out it was load-bearing when that person takes a fortnight off, and a task that took twenty minutes takes someone else a day and a half, or does not happen at all.
The audit, in about two hours
This part you can do yourself, and it is worth doing whether or not you ever speak to anyone about it. List every recurring task in the business that happens weekly or more often. Not projects — the repeating things. Then against each one write three things:
- Who does it, by name. If the answer is “whoever notices”, write that down too.
- What breaks if it is late — money, a deadline, a customer, or nothing.
- Whether anyone else could do it tomorrow from what is written down today.
The rows where column one is a single name, column two is “money”, and column three is “no” are your actual risk register. In a five-person company there are usually three to six of them. The owner can normally name two before writing anything down. The other three or four are the interesting ones.
Where it gets specific to you
Knowing which tasks are fragile is the easy half. Deciding what to do about each one depends on your numbers.
A weekly task that takes twenty minutes costs about seventeen hours a year. That is real, but it does not automatically justify building anything — if the build takes thirty hours and then needs maintaining, you have made your life worse and called it progress. The same task at daily frequency is a completely different decision. And a task that is cheap in hours but expensive when it goes wrong — a compliance deadline, a payment run — is worth systematising at almost any frequency, because you are not buying time back, you are buying the error rate down.
Which of your rows falls into which category depends on your volumes, your margins, and what a mistake actually costs you. That is the conversation, and it is not one anybody can have honestly from a blog post.
How we would walk you through it
We start from the list you already made — we do not ask you to redo it. Then we sit with whoever actually performs each task, because the owner’s version of a process and the real one differ more often than not. We come back with a short written map: what runs, what it depends on, where the single points of failure are, and a ranked view of which two or three are worth fixing first.
Which ones to leave alone is a real output, not a cop-out. Most audits produce a list of twenty improvements and deliver none of them. Three is a plan.
That mapping sits inside our digital and operations practice. Where the answer turns out to be a system rather than a process change, it hands over to workflow automation.
If you have made the list and want a second opinion on which rows matter, tell us what is on it — the three tasks you would least like to lose, and roughly how often each runs. That is enough for a useful first conversation.
A question worth answering
If the person who does your most fragile weekly task were unreachable for two weeks starting tomorrow, what specifically would you do on day one? If you cannot answer that in a sentence, you have found your first row.
