Month-End Close, Simplified

Every CFO has heard the pitch: cut your close from twelve days to five. The vendor demos look clean, the case studies sound believable, and the project gets approved. Eighteen months later the close is at ten days, half the controls are bypassed, and the team is exhausted. The compression happened in the wrong places.

This is the playbook we use when we're asked to actually shorten a close — meaning shorten it sustainably, with controls intact and a team that isn't on the brink. It's not glamorous and it doesn't lead with a tool. It works.

Diagnose before you compress

The close is a sequence of dependencies. You can't shorten it without knowing which step blocks which other step. Most teams have never mapped this. They have a close calendar, but the calendar reflects how the work happens to be sequenced, not how it could be.

Start with two weeks of diary-style tracking. Every accountant logs the work they actually do during close — task, hours, what they were waiting on. The output is a real dependency map. It almost always reveals two things: a handful of tasks blocking everything else, and a long tail of work that could move pre-close but doesn't.

The 80/20 of close time

In every close we've diagnosed, roughly 20% of the tasks consume 70–80% of the elapsed time. The usual suspects:

• Intercompany reconciliations — because nobody owns the difference until close week, when it explodes.
• Manual accruals — usually for the same recurring items, computed from scratch each month.
• Sub-ledger to GL tie-outs — done manually because nobody trusts the automated feed.
• The "review" step — controllers reviewing what they should have built into the prepare step.

Fixing those four areas typically takes a 10-day close to a 5-day close. Tools help, but the savings come from rethinking the work, not from installing the tool.

Continuous accounting isn't a product. It's a habit: doing in week three what you used to leave for week four.

The four moves that actually compress the close

1. Pre-close what doesn't need to wait

Recurring journals, depreciation, prepaid amortization, lease accounting — none of these need month-end data. They can run on day 25 with adjustments at month-end if needed. We've seen teams save 1–2 close days just by shifting these.

2. Drive sub-ledgers to "soft close" by day -1

AP and AR should be substantially closed before the GL close begins. That means cut-offs published in advance, late invoice policy enforced, and bank recs current. If you start day one of close still chasing AP, the close is already late.

3. Replace manual recs with exception-based recs

The point of a reconciliation isn't to recompute the balance. It's to identify the differences. Modern reconciliation tools (and well-built ERP automation) flag exceptions automatically — your team works the differences, not the matches. This alone often cuts 1–2 days.

4. Move the review left

If the controller is the first person to look critically at the numbers, you've built a review delay into the close. Build the analytical review into the preparer's workflow — variance flags, prior-period comparisons, materiality thresholds — so what reaches the controller is already curated.

What not to do

• Don't start with a tool selection. A tool deployed on top of a broken process amplifies the brokenness and gives leadership false confidence.
• Don't compress by skipping review. Speed without controls creates restatements. The savings have to come from making review faster, not from removing it.
• Don't measure only elapsed days. Add quality measures: late adjustments, audit findings, team overtime. A "5-day close" with three post-close adjustments isn't a 5-day close.
• Don't try to fix everything at once. Pick one entity, one cycle, get to the target. Then propagate. Trying to compress all entities simultaneously is how 18-month projects happen.

Where to start

The diagnostic comes first. Two weeks of close tracking and a half-day workshop with the close team will almost always reveal where 3–5 days of slack are hiding. From there, the work breaks into a 90-day plan: 30 days to pre-close and sub-ledger discipline, 30 days to rec automation, 30 days to review redesign. We've watched 12-day closes get to 6 days in a single quarter using exactly this path — no new ERP, no new tool, no new headcount.