The Month-End That Takes Two Weeks

The month closes on the 31st. The numbers arrive on the 14th.

By then the month they describe is half over, the decisions they would have informed have already been made, and everyone reads them as history rather than as information.

A number that arrives too late to change anything is a record, not a report.

Where the two weeks actually go

Rarely in the accounting. Almost always in the gathering.

Chasing what was never captured. Cash variances with no note, comps with no record, a night where the till was reset and nobody wrote down why. Each one is a phone call, and the calls are what take the time.

Reconciling things that should have matched. Card settlements against till totals, delivery notes against invoices. The work is not the reconciliation; it is finding the one that does not match and working out why, three weeks later, from memory.

Waiting for documents. Supplier invoices that arrive on the 8th, a statement that comes quarterly, an expense claim someone forgot.

One person doing all of it. Frequently the manager, in gaps between shifts.

The reframe that helps

Stop trying to make month-end faster. Move the work to the night it happens.

Almost everything above is easy on the night and hard three weeks later, because on the night the person who knows the answer is standing there.

a cash variance explained on the night → one line
the same variance in three weeks       → a phone call, a guess, or an unexplained figure

The cost of capture is roughly constant. The cost of recovery grows every day.

What to move to the night

Four things, none of which takes more than a minute:

The variance note. Whoever cashes up writes what they think caused it. Even "no idea" is useful — it distinguishes an unexplained variance from an unexamined one.

Comps, as they are given. Rung as comps rather than remembered.

Anything unusual, in one line. A till reset, a refund, a supplier paid in cash, a machine that went down. The line takes seconds and saves the phone call.

The deliveries that arrived. Signed against what was ordered, not against what the invoice says three weeks later.

What genuinely cannot be moved

Some of it is real waiting and no process will fix it.

Supplier invoices that arrive when they arrive. Quarterly statements. Anything that depends on a third party's calendar.

⚠️ For these the answer is not to speed them up. It is to produce a report that does not wait for them — a fast operational close with the known figures, and a slower financial close for the record.

Most venue decisions need the first and are held hostage by the second.

The two-close idea

Operational close, on the 1st or 2nd. Revenue, labour, comps, obvious cost of goods. Accurate to a few percent, available while the next month can still be changed. This is the one that gets used.

Financial close, when everything is in. Complete and reconciled. This is the one that matters for tax, for the accountant, and for anything requiring precision.

⚠️ Trying to make one report do both jobs is why the useful numbers arrive too late to be useful.

The test for whether it is working

Not how long the close takes.

"When we make a staffing or pricing decision, are we looking at last month or the month before?"

If the answer is "the month before", the close is too slow regardless of what the calendar says, and the operational figures are the ones to pull forward.

Three to hold

Days from month end to usable operational numbers. The number that determines whether reporting affects decisions at all.

Share of cash variances with a same-night note. The single biggest driver of chase time.

Number of items still open at close. If it is the same items every month, they are structural, not accidental.

Where the record has to sit

If the till is one system, the rota another, comps nowhere, and the notes on paper behind the bar, the gathering is the work — and gathering is what takes two weeks.

tasteck keeps sales, shifts, and comps on the same nights, so the operational close is a report rather than a fortnight of assembly.

The month did not take two weeks to close. It took two weeks to find.