The year-end scramble is not caused by December. It is caused by the previous eleven months, and December is just where the bill arrives.
November is the last month in which you can do anything about it. By the second week of December the people you need answers from are unavailable, and the work stops being a close and starts being an archaeology project.
Where the time actually goes
Ask anyone who has done a year-end close in a small business where the hours went, and the answer is almost never “adding things up”. It is chasing: a missing invoice, an unexplained transfer, a contractor’s details, which of two similar payments was the deposit. Each item is small. There are forty of them.
Every one of those forty is a piece of context that existed at some point during the year and was not written down while it was cheap to write down.
The November pass
You cannot retrofit a year of record-keeping in November. You can do the three things that shrink the scramble most:
- Close the open questions while people still answer email. Go through the year’s unexplained items now, not on 28 December. This is the single biggest saving available and it is pure calendar arbitrage.
- Confirm anything that renews in January. Insurance, licences, subscriptions, leases. Renewals discovered late are expensive twice — the price and the panic.
- Write down what you had to reconstruct. A list of the things you could not answer this year is the specification for not having the problem next year. Nobody ever writes this list, which is why the scramble is identical annually.
That third one costs about twenty minutes and is the only item that compounds.
Where it depends on your situation
How much to systematise depends on how many places your data lives and how many people touch it. A single-owner business with one bank account and one card can run this on a checklist indefinitely, and building anything is over-engineering. Add a second entity, a property or two, a couple of staff with cards, and a payment processor that settles in batches, and the reconstruction cost rises faster than the transaction count — because the work is in the joins, not the rows.
Where your business sits on that curve determines whether the answer is a better habit or a better system. It is a genuine fork, and we have recommended both.
How we would walk you through it
We start with the reconstruction list — either the one you wrote, or the one we build by asking what took longest last year. Then we trace two or three of those items back to the moment the context was lost. It is nearly always a specific handoff: a payment made without a reference, a job closed without a note, an approval given verbally.
Fixing the handoff is usually small. Finding it is the work, and it is much easier from the outside.
This sits across CAO Digital & Operations and, where the fix is a build, workflow automation.
Tell us the one thing you always have to chase in December. That single item usually points straight at the handoff worth fixing.
A question worth answering
What did you have to reconstruct last year that you could have simply recorded at the time? If that answer is the same as the year before, the close is not the problem.
