Closing month-end in three days

A short month-end is not achieved at month-end. It is achieved in weeks one to three, by gating invoice readiness while the work is still happening.

Firms that close quickly are not working harder in the final week. They are releasing a position that was already assembled, rather than rebuilding one from scraps. Close becomes a release rather than a reconstruction, and the difference is entirely upstream.

This post is about the mechanics of that: what gets gated, when, and what a firm actually does on the days it currently does nothing.

One customer reported cutting their close from nearly two weeks to under three days after putting these gates in place. That is their own measurement of their own firm, not a projection of what yours would see — what moved for them was the amount of reconstruction left to do once the month ended.

What close is actually doing

Watch a slow close and you will see finance performing four distinct jobs that look like one.

Establishing what happened — chasing timesheets, confirming costs, working out which milestones were reached. Establishing what may be billed — applying rate cards, caps, discounts and variations that were agreed elsewhere. Resolving disagreement — the conversations about whether an overrun is billable, whether a variation was approved, whether a client will accept a line. And finally, producing the invoice.

Only the fourth of those is invoicing. The first three are investigation, and they are being done in the narrowest window of the month by the function with the least context.

Each of the first three can be done earlier. None of them has to wait for the calendar.

Week one: gate the inputs

The first gate is the least popular and the highest yield: time and cost entered and approved weekly, not monthly.

Weekly approval is not about accuracy for its own sake. It is about who is doing the remembering. A project manager approving Tuesday’s entries on Friday can tell you whether that work was in scope. The same person approving five weeks of entries on the third of the following month is guessing, and their guess becomes a WIP balance.

The gate has to be a state on the record, not a reminder email. Unapproved time is not eligible for WIP. That single rule converts a soft expectation into a visible blockage, which is the point.

Week two: gate the commercial changes

The second gate is change control, and it is where most of the month-end argument originates.

Any work outside agreed scope requires a priced, approved variation attached to the engagement before the work proceeds. Not documented afterwards. Not agreed verbally and written up when it becomes contentious.

In practice this means someone reviews, mid-month, every engagement where actuals are running ahead of the budget line — at task level, not contract level, because a contract-level view will not show you an overrun until it is too large to price gracefully. Each of those gets one of three outcomes: a variation raised, a decision to absorb it deliberately, or a correction because the time was booked to the wrong place.

All three are fine. The failure mode is the fourth outcome, which is nobody looking until the invoice is drafted.

Week three: gate invoice readiness

The third gate is a formal readiness state on each engagement, tested before the period closes rather than during it.

An engagement is ready when four things are true: all time and cost for the period are approved, every milestone has a state that matches reality, every variation raised in the period is approved or explicitly rejected, and the billing rules that will apply are the ones recorded on the engagement rather than ones somebody intends to apply.

Run that test in week three and the output is a list. Some engagements pass. The ones that fail name the specific missing item and the specific person who owns it, with a week left to resolve it.

That list is the whole intervention. It converts month-end from a search into a queue with owners.

The thirty-second test

There is a quick way to tell whether the gates are real. Take any invoice line and try to explain it back to its source — the work it relates to, the approval that cleared it, the contract term that permits billing it — in under thirty seconds.

If that takes half a minute, close is a release. If it takes a phone call, close is an investigation, regardless of how quickly the invoices eventually go out.

What makes the test passable is that delivery evidence and commercial rules sit on the same governed record from the point of sale, so the invoice is assembled from entries that each already carry their justification. Our invoicing and WIP page follows a single line from timesheet to invoice.

Why the first two cycles feel worse

Gating readiness does not remove friction. It moves it earlier, and earlier friction is more visible.

In cycle one, work that used to be quietly absorbed at close now surfaces as a blocked invoice in week two, attached to a named person. Project managers who were comfortable approving time in arrears find themselves doing it weekly. Account leads who preferred to sort variations out at the end are asked to raise them in the moment, when the client conversation is harder because the work has not been done yet.

The complaints in that first cycle are legitimate. The firm genuinely is doing more work in weeks one to three, and it has not yet banked the saving at close.

Firms that cannot hold the line through two cycles will not get the benefit, and will conclude the discipline does not work. That is the most common failure of this change and it is a management problem, not a systems one. It is worth deciding in advance who will hold it and what they are permitted to say no to.

Boring is the objective

A good month-end is uneventful. Nothing is discovered, because everything was already known; nothing is argued, because the arguments happened when the facts were fresh and cheap to establish.

That is a lower ambition than it sounds and a harder one to reach. It does not require finance to work faster. It requires delivery and sales to close their positions on time, every week, when nothing is yet on fire — which is exactly when the incentive to do it is weakest. Where those positions get set is covered on our scoping and quoting page, and what it looks like in a technology firm on the IT services and consulting page.