Crestpoint brought month-end close down from 19 days to 8.
Crestpoint's close checklist ran over 100 tasks, most of it sequential, and the first several days were spent clearing exceptions and reconciliation gaps that had accumulated over the month. The CFO didn't just want a faster close — he wanted close to stop being the reason the team couldn't do forward-looking work the other three weeks of the month.
Close AgentThe bottleneck
Close consistently ran 17 to 21 business days, well past the 10-day target the CFO had set two years earlier without ever hitting it.
Flux analysis for anything outside threshold was written from scratch every period, usually by the same senior accountant who also owned three other close-critical tasks, making her the bottleneck for the whole checklist.
Recurring journal entries for accruals were rebuilt manually each month even though the underlying logic almost never changed — same vendors, same allocation methodology, different numbers.
Reconciliations were treated as a close-week activity rather than a monthly one, so every bank and balance-sheet account effectively had a month's worth of unreviewed activity waiting on day one.
The workflow
The Close Agent now runs reconciliation pre-checks throughout the month instead of waiting for close week, so the checklist starts from a much shorter list of open items on day one.
It pre-drafts recurring accrual entries against the prior period's methodology and flags anything that deviates from pattern — a new vendor, an allocation that moved — for review instead of rebuild.
It generates first-pass flux commentary by comparing current period to prior period and budget, pulling the underlying driver (volume, rate, timing) rather than just the variance number, so the accountant is editing, not drafting.
The controller's team reviews and approves rather than building each piece from a blank page, and the close checklist tracks progress against the same project tool the team already used, so nothing changed about how status got reported up.
“Eight days doesn't sound dramatic until you remember we spent two years stuck at nineteen. The team didn't get bigger, the backlog just stopped building up.”
VP of Finance, Crestpoint Manufacturing
Lessons learned
Close speed is a monthly problem wearing a close-week costume
Crestpoint's 19-day close wasn't caused by close week — it was caused by a month of unreviewed activity landing all at once on day one. Moving reconciliation earlier in the month did more for the timeline than anything done during close itself.
Automate the task your best person shouldn't be doing
The senior accountant rebuilding accruals from scratch every month wasn't a skills gap, it was a triage failure — the most valuable person on the team was doing the most repeatable work. Look for that mismatch first; it's usually the highest-leverage fix available.
Keep the reporting surface the same
The Close Agent plugged into the checklist tool the team already used to report status to the CFO. Nobody had to learn a new system to see progress, which meant leadership adopted the change without anyone having to sell it to them separately.
Draft-and-review beats build-from-scratch for buy-in
Giving accountants a first-pass flux narrative to edit, rather than asking them to trust a fully automated number, kept humans in the loop on judgment calls while still eliminating the blank-page time. That framing made adoption a non-issue.
“My best accountant used to spend the first week of every close on things a template could do. Now she spends it reviewing judgment calls, which is the part I actually need her for.”
Controller, Crestpoint Manufacturing