What a Monthly Close Actually Looks Like on a Legacy GL

It's 7:12 AM on the third business day of the month. Monica, Director of Finance at a five-site behavioral-health nonprofit with a $34 million operating budget, is 90 minutes into her day and staring at a Citrix session window. The Citrix session is Dynamics GP. The GP screen is running a trial-balance export for one of the five sites. It has been running for eleven minutes. She has four more sites to go.

Monica knows this exact keystroke sequence. She has run it every month for six years. Log into Citrix. Wait for the tunnel. Launch the GP client. Pick the company. Pick the fiscal period. Change the account range. Change the sort. Export to Excel. Save to the shared drive using the file-name convention she and her Controller agreed on in 2019. Repeat.

Today the export will run five times for five entities. Tomorrow she will do it again for six divisions inside one of those entities, then for the twelve program cost centers restricted-fund reporting requires. By Friday she will have 47 CSV files in a folder called `Close-2026-08`, and she will have not opened her budget model once.

The 2026 budget her team spent nine weeks on in September and October is sitting untouched. The rolling forecast the board asked for at the last meeting is sitting untouched. The variance narrative her CFO wants by Day 8 is sitting untouched. Everything downstream of the close is waiting for the close, and the close is waiting for the exports.

This piece is written for the Controller, the FP&A Manager, and the VP Finance who know this rhythm and who are quietly aware that most of the effort of the month is not the accounting. It is the assembly. It is the four-to-eight days of pulling numbers out of the ledger by hand so that the rest of the month can happen. And it is the shadow layer of Excel that grew around the ledger to do everything the ledger cannot do on its own.

‍

The first three days of the close

Every legacy-GL close has the same opening shape. Different clients, different vertical, different terminology, same pattern.

The Controller and the closing-side accountants finish period-end postings, run accruals, close the sub-ledgers, and mark the fiscal period closed inside the GL. That work is the closing itself, and on most teams it takes two to three days after month-end. It is the part the ledger was designed for. Dynamics GP, Blackbaud Financial Edge NXT, MIP, SYSPRO, Multiview, and the other on-prem systems all handle it well. They post transactions, calculate depreciation, run allocations, roll the sub-ledgers, and give you a closed trial balance you can trust.

Then the ledger's job ends. And the FP&A team's job begins.

The FP&A team's job is to turn the closed trial balance into something people can read. A management P&L by department. A cash-flow variance narrative. A rolling forecast that reflects what the actuals just told us. A board pack that shows the CFO's story of the month. A grant-restricted report for the federal funder. A wage-mix variance for the CNO or the plant manager. A YTD-vs-plan table that the CFO can walk through in Monday's leadership meeting.

None of that comes out of the GL natively. The trial balance comes out of the GL. The management story does not. Between the closed trial balance and the management story is the assembly step, and the assembly step is where the days go.

‍

The assembly step, mechanically

Assembly starts with the export. On Dynamics GP, a Controller runs Smart Lists or Management Reporter, exports a trial balance and a divisional P&L, and drops them into a shared drive. On Blackbaud Financial Edge NXT, she runs Financial Statements or the FE NXT report writer, exports to Excel or PDF, and does the same. On MIP, she runs a series of Statement of Financial Position and Statement of Activities reports, one per fund, one per program, and exports each. On SYSPRO or Sage 100, she runs the equivalent GL reports out of the desktop client. Steeve Wintle, Controller at Adhesive Technologies, put the pre-Centage version of this shape plainly in a customer conversation: "SYSPRO ERP could not produce financial statements natively. Building a balance sheet or income statement required manual workarounds and hours of Excel gymnastics."

‍

The exports land as CSV or XLSX files with the ledger's account numbers, the ledger's chart, the ledger's period stamps, and the ledger's rounding. Every file is a fresh snapshot of a slice of the trial balance. The FP&A team then has to do six things with those files before the numbers become useful.

1. Consolidate. Multi-entity or multi-site organizations have to bring the exports together. For a five-site nonprofit, that is five CSVs into one workbook, with intercompany eliminations run by hand. For a three-plant manufacturer, that is three site exports plus a corporate consolidation entry. For a multi-fund government or nonprofit organization, that is one export per fund plus a set of inter-fund transfers. Every consolidation is a Vlookup or Index-Match chain, and every consolidation is a place where the file structure can silently change.

2. Re-map. The GL's chart of accounts is the ledger's chart, not the management chart. The board sees revenue by program, not revenue by GL account code. A management P&L rolls thirty accounts into "Salaries and Wages" and forty accounts into "Program Delivery." That rollup is a mapping table that lives inside Excel. The mapping is fine as long as nobody added a new GL account this month. If somebody did, the mapping quietly misses it and the total is quietly wrong.

3. Layer in the sub-ledger detail. GL summary is not enough for the board pack. FP&A wants payroll detail by position, capex detail by project, grant spend by grant, unit sales by SKU. The sub-ledger sits in a different report or a different system. That report gets exported to a different CSV that gets joined to the GL export inside the master workbook.

4. Compare to budget. Once the actuals are consolidated and re-mapped, they get compared to the current-year budget. The budget lives in the Excel model the team built in September and October. The comparison is a set of formulas that assume a specific line-item layout in the budget tab. If the layout is right, variance falls out. If the layout drifted, variance is off by a row and every commentary is wrong.

5. Roll the forecast. After variance is calculated, the rolling forecast updates. The rolling forecast is a separate tab, or a separate workbook, that pulls current-month actuals and re-projects the remainder of the year. On a driver-based model, the driver assumptions flex on actuals. On a plug-model, the analyst goes in and adjusts by hand.

6. Build the board pack. The board pack is a set of tabs. Executive summary. Consolidated P&L. Cash flow. YTD vs plan. Prior-year comparison. Program or plant detail. Workforce roll. Grant-restricted rollups. Each tab pulls from the consolidated actuals tab, and each tab has its own layout the CFO signed off on last quarter.

‍

Every one of those six steps is a place where a formula can silently break. Every one of them takes hours or days of a finance analyst's time. And every one of them repeats every month.

Mary Flannery, Director of Finance at God's Pantry Food Bank, described the pre-Centage version of this shape in a customer conversation. Her verbatim:

‍"We'd spend days just cleaning up formulas. If something changed mid-year, it meant hours of rework."

The read: the days-just-cleaning-up-formulas time is not close work. It is assembly work. It happens after the ledger is closed. It happens because the assembly lives in Excel, and Excel is a chain of anchored formulas that breaks silently when the chart, the account list, the mapping, or the layout drifts. Days of cleanup is a normal month on a legacy-GL FP&A workflow. It is not a sign of a bad team. It is a sign of an assembly step that has no infrastructure of its own.

‍

The Excel shadow layer

Every legacy-GL FP&A team has one of these. Sometimes it is polite. A single master workbook, ten tabs, a set of clear mapping tables, a version-control convention that mostly works. Sometimes it is not polite. Nineteen linked workbooks, three of which are on the CFO's laptop and cannot be opened by anyone else, connected by external references, formula chains that go across files, one broken link away from a Sunday-afternoon phone call.

This is the Excel shadow-FP&A layer, and it grows around every legacy GL for the same reason. The ledger holds the numbers. The ledger does not hold the plan. The ledger does not hold the workforce model. The ledger does not hold the driver assumptions. The ledger does not hold the board pack. Somewhere between the ledger and the board, all of that has to live. On most legacy-GL FP&A teams, "somewhere" is Excel.

The shadow layer takes different shapes depending on the vertical.

At a mid-market manufacturer running Dynamics GP or Sage 100 or SYSPRO, the shadow layer typically includes a master consolidation workbook, a per-plant cost workbook, a workforce roster the HR team keeps and the finance team receives once a quarter, a capex tracker, and a rolling forecast the CFO uses in weekly leadership meetings. Five workbooks. Fifteen formula chains between them. One person on the team who knows how they connect.

At a mid-sized nonprofit running Blackbaud Financial Edge NXT or MIP or Multiview, the shadow layer typically includes a fund-allocation workbook, a grant-by-grant restricted-spend tracker, a position-level personnel budget the HR team maintains, a program cost-per-participant model the ED uses for board reporting, and a multi-fund cash-flow model that the Director of Finance rebuilds every quarter because the fund structure keeps changing. Five workbooks. More formula chains than that. Personnel planning specifically is a known gap on these systems: at one nonprofit museum on Financial Edge, the operator's shorthand is that personnel planning is entirely manual in spreadsheets, because the ledger has no workforce planning capability of its own.

At a mid-sized professional services firm running Deltek or Elite, the shadow layer typically includes a time-and-billing rollup workbook, a project-profitability workbook, a partner-compensation workbook, a workforce utilization model, and a rolling revenue forecast keyed to backlog and hires.

Vertical to vertical, the specifics change. The pattern does not. The ledger holds the actuals. Excel holds everything else. Every month, actuals get pulled from the ledger, dropped into Excel, and reconciled against the everything-else. And every month, the reconciliation is where the assembly hours go.

‍

Where the hours go, honestly

A useful exercise: count the hours.

The internal evidence we work from, from customer conversations and FP&A-team interviews, points at a range. On the light end, a two-to-three-person FP&A team at a mid-market company on Dynamics GP or Sage 100 or Blackbaud FE NXT can compress the assembly step into three to four full days a month, and total post-close assembly runs 40 to 55 hours across the team. On the heavy end, the same team can absorb 60 to 70 hours a month in assembly work if the ledger is behind, the chart changed mid-year, or the CFO wants extra cuts on the board pack. Multiply by twelve months and you get the pattern one operator at a fund-restricted nonprofit described as roughly 600 hours a year gone to the assembly step alone. That is not the close. That is what happens after the close, before anything strategic can start.

Another anonymized pattern we see, at a mid-sized restricted-funding nonprofit: nineteen linked workbooks live inside the FP&A layer, one per fund and one per program. Each workbook pulls from the master consolidation tab, applies its own allocation math, generates its own board view, and reports out. The Director of Finance can trace exactly how each workbook works. Nobody else on the team can. When the Director of Finance takes a vacation, the close waits.

Neither pattern is a sign of a broken team. Both are the natural consequence of a legacy ledger that holds numbers well and does nothing else, combined with an FP&A team that needed planning, workforce, allocation, and reporting capability the ledger could not provide. Excel was the only tool with enough flexibility to hold all of it. So the team built the shadow layer, and the shadow layer became the FP&A infrastructure.

The cost is not a lack of capability. The FP&A team is capable. The cost is where their hours go. Sixteen to twenty-four days of finance-team time each month sitting inside the assembly step is time not sitting inside the strategic step. It is the number of variance conversations the CFO does not have with the plant manager. It is the number of program-cost analyses the ED does not run before pricing next year's grant proposal. It is the two weeks between month-end and board-ready that the audit committee reads as "your team is behind" and the team reads as "we are running as fast as we can."

The audit committee is not wrong. The team is not wrong either. The infrastructure is wrong.

‍

What the workflow looks like when the dumps go away

The mechanical fix is not more Excel. The mechanical fix is a connector that pulls the actuals from the ledger on a schedule, an FP&A system that holds the mapping, the plan, the workforce model, the driver logic, and the board pack in one place, and a workflow where the FP&A team never runs the export step by hand.

That is what a connected FP&A layer does. The ledger stays where it is. Dynamics GP stays on-prem. Blackbaud Financial Edge NXT stays wherever Blackbaud runs it. MIP stays on the finance-team desktop or on Community Brands' hosted environment. The connector runs on a schedule, pulls the trial balance and the sub-ledger detail into the FP&A system, translates it through a mapping table that lives in the FP&A system, and drops it into the FP&A model.

The FP&A team's assembly step disappears because the FP&A system holds the pieces that used to live in nineteen workbooks. One chart-of-accounts mapping. One workforce model. One driver-based forecast. One consolidation logic. One board pack layout. Actuals flow into the model on schedule. Variance calculates on the model. The board pack renders from the model.

‍

Steeve Wintle, Controller, Adhesive Technologies. Adhesive Technologies is a specialty-chemistry manufacturer running SYSPRO on-prem, a legacy ERP that gives a Controller the ledger she needs and none of the financial-statement tooling. Wintle's verbatim on what the connected workflow feels like from inside his team:

‍"Centage like auto downloads and uploads from SYSPRO out to whatever a CSV or whatever it is. And then uploads into Centage automatically. I never have to think about it."

The read: "I never have to think about it" is the operative phrase. His team is not thinking about the Citrix session, the trial-balance export, the file-naming convention, the shared-drive folder, the Vlookup chain, or the mapping table. They are thinking about the plant. Because the assembly step is running on a schedule inside the FP&A layer, the mental load of the close moves off the actuals-pull and onto the analysis.

‍

Mary Flannery, Director of Finance, God's Pantry Food Bank. God's Pantry is a regional food bank on Dynamics NAV. NAV has a native Centage connector, meaning actuals flow automatically. Flannery's verbatim on what changed:

‍"We used to spend days cleaning up formulas. Now the numbers just show up."

The read: "just show up" is the difference between an assembly workflow and a connected workflow. The team's hours moved off the assembly step. What lands on the model is the ledger's actual actuals, translated through a mapping that lives in one place, refreshed on a schedule that the team does not have to babysit.

‍

Neither example is a claim that the ledger got easier. SYSPRO is still SYSPRO. NAV is still NAV. What changed is where the FP&A model lives. It moved off the workbook chain that was glued to the ledger and into a system that holds the model, the mapping, the workforce roster, and the board pack together.

If you want the specific technical question of which connectors are truly native, which are scheduled batch, and which are file-import, and how to interrogate any FP&A vendor about your specific GL, that is what our connector-that-holds blog covers. It is the mechanical companion to this piece. It matters because "connected FP&A layer" is not one thing. It is a tier stack. Different GLs sit at different tiers, and the question a Controller should ask on the demo is which tier her ledger sits at.

If you want the factual read on what Dynamics GP's and QuickBooks Desktop's lifecycle dates actually mean for a working close, whether you stay or go, that is the Great Plains and QuickBooks Desktop end-of-life blog. No countdown clock. Just the dates and what they change about the close.

‍

What to check in your own team

If you are running the close on a legacy GL today and want to know whether your Excel shadow layer is in the range where a connected FP&A layer would return meaningful hours, a short internal check helps.

Count the number of workbooks the FP&A team opens between Day 3 and Day 10 of the close. If it is more than five, the shadow layer is doing work the ledger cannot do, and that work is worth naming.

Count the number of exports the Controller or the Senior Analyst runs by hand from the GL each month. If it is more than four, the assembly step is where the hours are going.

Count the number of formulas across those workbooks that broke or needed cleanup in the last twelve months. If your team can name specific breakages, that is not a talent gap. That is the shadow layer signaling that a formula chain is holding up a mission-critical process.

Count the number of days between month-end and the CFO's variance narrative being board-ready. If it is more than seven, the assembly step is dominating the close.

Finally, ask the Director of Finance or the FP&A Manager how much of last month's calendar went to assembly versus how much went to analysis. If the answer is more than fifty percent assembly, the team is running a legacy-ledger workflow with no FP&A infrastructure of its own, and the hours are absorbed by people rather than by tools.

None of those numbers are a judgment. They are a diagnostic. The Excel shadow layer grew for real reasons: the ledger could not do planning, could not do workforce, could not do driver-based forecasting, could not do fund allocation, could not do the board pack. Excel could hold all of it, so the team put all of it there. The question is not whether the shadow layer works. The question is what the team's hours would look like if the assembly step ran on infrastructure the FP&A team owned instead of on formulas the FP&A team maintained.

‍

Take-home

A legacy-GL close is two workflows, not one. The ledger's workflow is the close itself, and on-prem systems handle that piece well. The FP&A team's workflow is the assembly step that turns the closed trial balance into the plan, the workforce roll, the variance narrative, and the board pack. On most legacy-GL teams, the assembly step lives in Excel, and Excel is where the days of the month go.

The fix is not a new ledger. The fix is a connected FP&A layer that runs the assembly step on a schedule, holds the mapping, the workforce model, and the board pack in one place, and gives the FP&A team back the sixteen-to-twenty-four days a month that currently go to the exports and the reconciliation.

If your Day 3 mornings look like the Citrix session above, price it before you fix it. Get the Legacy-GL FP&A Integration Scorecard — its close-hours calculator takes the four phases you just read and puts your own hours and loaded cost against them. Most teams are surprised by the annual number.

‍

If you would rather walk it through with someone, book a 15-minute close diagnostic with a Centage FP&A operator who has run a close on Dynamics GP, Blackbaud Financial Edge NXT, or MIP. Bring your close calendar.

Centage is FP&A software built for finance teams at growing mid-market companies on legacy or on-prem GLs. The workforce module, the driver-based forecasting, the multi-entity consolidation, and the board reporting sit in one connected system, with supported connectors into Dynamics GP, Sage 100, Blackbaud Financial Edge NXT, Dynamics NAV, SYSPRO, Elite, Multiview, and the modern cloud ledgers. Implementation is four to six weeks, finance-owned, no IT project. When actuals flow on schedule, the assembly step goes away, and the FP&A team goes back to being an FP&A team.