← ALL CASE STUDIES
Case Study · Finance & Accounting

Month-end close, three days early.

Industry benchmarks put the average close at 6–10 business days — and most of that is not accounting. It is waiting: for statements, for receipts, for the client to answer one question. This bookkeeping firm automated the waiting. The accounting got its month back.

Day 3Close day, was day 10
94%Transactions auto-matched
0Clients blocking close unseen
Bookkeeping & CAS firm
IndustryFinance & Accounting
Clients60+ monthly bookkeeping clients
Built byPromata — Blueprint → AI Build
StatusLive · closing right now

Does this sound like your firm?

The close is a controlled fire drill — checklists, email threads, and a spreadsheet tracker held together by heroics.
The books could start on the 1st. They start when the last statement shows up.
One missing account holds the whole file hostage.
Your seniors spend close week on cleanup and chasing, not review.
Advisory conversations that should happen on the 10th happen on the 20th — after the client already decided.
IF YOU CHECKED THREE OF THESE — THIS CASE STUDY IS ABOUT YOU.
The Challenge

The close was not slow. The waiting was.

Across sixty clients, the same sequence repeated monthly: request documents, wait, remind, wait, reconcile in a rush, query the client, wait again. The preparation and review took the same hours every month — the calendar damage came entirely from the gaps between them.

⚠ CLOSE WEEK, TRADITIONAL
Day 4: still waiting on 22 of 60 clients’ statementsBOOKS UNTOUCHED
“Just checking in!” — email #3 to the same clientSTAFF AS REMINDER APP
Uncategorized pile queried in one giant list, lateCLIENT OVERWHELMED, SILENT
Senior accountant re-keying statement PDFsCLEANUP, NOT REVIEW
Advisory call pushed to the 20thDECISIONS ALREADY MADE
✦ CLOSE WEEK, AUTOMATED
Statements requested day −5, chased automatically, validated on arrivalIN BY DAY 0
Reminders escalate politely until the file landsZERO STAFF TOUCHES
Queries sent one at a time, in context, same dayANSWERED IN HOURS
94% of transactions matched before a human looksREVIEW, NOT CLEANUP
Advisory call on the 5th, with fresh numbersTHE FIRM IN THE ROOM
VS

THE SAME SIXTY CLIENTS — A FIRE DRILL VS. A CLOSE THAT RUNS ITSELF

  • Waiting was the product defect

    Clients do not judge the quality of a reconciliation; they judge when the numbers arrive. Every waiting day pushed the firm from advisor to historian.

  • Chasing consumed senior hours

    Reminder emails, status checks, and re-requests are clerical — but they were done by the people qualified to review, because only they knew what was missing.

  • The bottleneck compounded across clients

    Sixty closes sharing one team means one late client steals attention from fifty-nine punctual ones. Fragmentation, not volume, was the killer.

  • Month-end set the firm’s ceiling

    Capacity for new clients was capped not by accounting skill but by how much chasing the team could absorb.

The Solution

The close starts before the month ends.

Five days before month-end, the system requests every client’s documents — then chases, validates, matches, and queries continuously, so that by day 1 the books are largely reconciled and humans review instead of assemble.

1
2
3
4
5
6
7
8
9
10
▲ CLOSE NOW: DAY 3CLOSE BEFORE: DAY 10
  • Collection runs on the calendar, not on memory

    Requests go out day −5 with exactly what is needed; reminders escalate on a cadence; arrivals are validated for date ranges and completeness the moment they land. The receipt phase compressed from weeks to days.

  • Matching before humans

    Bank and card transactions auto-match against rules learned from the firm’s own history — 94% resolve untouched, and the rest arrive grouped and contextualized for review.

  • Queries in context, one at a time

    Instead of a month-end wall of questions, clients get small, specific asks while the transaction is fresh — answered in hours, not weeks.

  • A dashboard instead of status meetings

    Partners see every client’s close state live — who is blocking, what is missing, what closes today. “Where are we on Northway?” stopped being a meeting.

Results & Business Impact

The 10th became the 3rd.

Day 3Median close, was day 10
94%Transactions auto-matched
−7Days of waiting removed
+40%Client capacity, same team
  • Seven days of waiting, deleted

    Preparation and review hours stayed constant — the calendar time between them collapsed. That is exactly what the research on close cycles predicts, and exactly what happened.

  • Advisory became the product

    With numbers fresh by the 5th, month-end conversations moved from “here is what happened” to “here is what to do” — the work clients actually pay premium fees for.

  • Capacity without hiring

    The chasing hours — the ones that scaled with every new client — are gone. The firm onboarded new monthly clients with the same headcount.

  • Close week stopped being a personality test

    No heroics, no fire drill, no burned-out seniors. The checklist runs; the humans judge.

The books were never the bottleneck. The inbox was.

Accounting firms do not lose days to debits and credits — they lose them to waiting for the things debits and credits need. Automate the waiting, and the same team closes faster, serves more clients, and finally does the advisory work it was built for.

Day 3Median close
94%Auto-matched
+40%Capacity, same team

What we didn't solve

The close assembles itself; the judgements inside it do not. Reconciliation exceptions are surfaced with the discrepancy shown and left for an accountant to decide — the system never picks a side automatically, because a silently resolved mismatch in a ledger is worse than a slow one.

What day does YOUR close actually land?

Tell us your client count and your close day. We’ll show you where the waiting hides — before you spend a dollar.

Book a discovery call

Want results like these?

Tell us where your team is losing hours. We'll show you exactly what automation can do about it — and what it's worth, before you spend a dollar.

30 minutes · No pitch deck · An honest first read

Case study: month-end close, three days early