Rebuilding a $50M contractor's back office in three months
A pre-construction and restoration firm ran its entire operation by hand, until a single bad project forced a better way. Folio rebuilt it without replacing the system the team trusted.
Pre-Construction & Restoration · Ontario, Canada
- To build and go live (vs. 6–18 mo industry norm)
- 3 months
- Month-end close
- 10 → 3 days
- Less time spent on AP
- ~80%
- Of admin returned to the team
- 50+ hrs/mo
To build and go live (vs. 6–18 mo industry norm)
Month-end close
Less time spent on AP
Of admin returned to the team
The Company
A roughly $50M Ontario-based pre-construction and restoration contractor running multiple concurrent projects on a custom, in-house accounting system, built years prior. Their system was the financial source of truth that the entire company depended on, but it sat at the center of an otherwise fully manual operation: no automation, no modern ERP, no real-time visibility around it.
The Challenge
The back office ran on paper and effort. Invoices were either dropped off in person, or sent as multi-page PDF packets which were keyed in by hand, roughly 400 a month. PMs hand-coded every cost line. Nothing linked invoices to POs automatically, there was no threshold-based approval routing, and field supervisors couldn't issue a PO without routing through the office. Job financial health was only visible after the fact. Budget variances were surfaced at month-end, long after the moment to act had passed. Month-end close itself was a manual reconciliation grind that stretched around ten business days.
None of that was visible as a single problem from the inside. What leadership saw was a high-profile municipal project that went sideways, and a growing sense that the business was working harder than it should have to. They knew something was wrong. They didn't yet know what. So they went looking for outside help to make the operation more efficient and take error out of the process.
The Diagnosis: Folio's Audit
Folio didn't start by building anything. It started by listening. The engagement opened with a full audit of the company's internal systems and workflows: how invoices actually moved, where PMs spent their time, how a job's financial health was tracked, and where the legacy accounting system held the business together versus quietly created work.
The numbers told the story leadership couldn't see from inside the day-to-day. Roughly 280 hours a month, close to two full-time roles, were going to manual back-office work that didn't need a human: hand-keying some 400 invoices, manually coding every cost line (about one in ten miscoded and caught only at close), chasing field POs through the office, and grinding through a ten-day reconciliation each month. All in, roughly $150,000 a year in labour was tied up in work that should have been automated.
The more dangerous finding was on the jobs themselves. Budget variances were surfacing four to six weeks late, at or after month-end, long past the point of correction. Across active projects, the audit identified roughly $500,000 in cost overruns over the prior year that had gone unflagged until it was too late to act. The municipal project hadn't failed because of one bad decision. It failed because the business had no instrument panel, and the same blind spot was quietly costing margin on jobs no one was worried about yet.
That finding shaped everything Folio built next, and it's why nothing got ripped out that didn't need to be.
What Folio Built
The obvious move would have been to rip out the legacy system and force a new off-the-shelf ERP on the team. Folio did the opposite, and that decision is why it worked. The central decision wasn't technical, it was architectural: rather than migrate decades of financial history off the legacy system, the path most likely to break trust and stall adoption, Folio kept it as the system of record and built an automation and reconciliation layer on top, syncing cost data in both directions so the books never drifted. Where a generic ERP forces a team onto the vendor's process and stalls on adoption, a system built to fit meant the rollout was a non-event rather than a fight. Over a single three-month engagement, that meant:
- AP, fully automated
- OCR for physical invoices via a text message, batch PDF packets auto-split into individual invoices, invoices auto-linked to PO numbers on import, a sequential assignment workflow that ended manual field-filling, and automated month-end alerts for pending invoice and labour allocations.
- Cost coding without the keying
- Codes applied automatically when a PM selects a budget line, and automatic project creation detected from job-cost historical uploads.
- POs from the field
- Text and voice-memo PO creation for supervisors on mobile, with large commitments ($100k+) auto-routed to the PM for approval.
- A job-health early-warning system
- Automated variance alerts at 5% and 20% thresholds, with logic that learns from how users respond and escalates to the right stakeholder, plus auto-generated monthly PM performance reports.
- Automated reconciliation and bid packages
- Monthly cost balancing back to the existing accounting system, and automated bid-package creation built for an estimating-heavy pre-construction shop.
And Folio stayed through support. This was never a hands-off-the-deck engagement.
Results
Built and live in three months: against an industry-standard 6–18 months for an ERP rollout at this scale, and an enterprise-software average closer to 21. A system the company owns outright, with no per-seat licensing scaling against headcount as it grows.
Full adoption, office and field: because the system fit existing workflows instead of replacing them, the team actually used it, sidestepping the stalled, partial rollouts that sink most ERP projects.
Manual back-office work collapsed: the bulk of that ~280 hours a month eliminated across AP, cost coding, PO handling, and reconciliation, with AP processing time alone down ~80%. Capacity the company would otherwise have had to hire for as it grew.
Coding errors all but gone: PMs no longer hand-code cost lines, eliminating the cleanup that used to eat into every close.
The blind spot, closed: the ~$500K-a-year exposure the audit surfaced is exactly what the new early-warning system is built to eliminate. Variances now surface automatically at 5% and 20% thresholds, weeks before they'd previously appear. The overrun pattern that first brought Folio in is now caught while there's still time to act.
Month-end close cut from ~10 business days to 3, reconciled automatically against the system of record.
Senior staff off admin, onto margin: estimators and operations leads moved from administrative cleanup to higher-value work.
“I came into the business knowing what good operational processes should look like, but the reality was that many of our systems and workflows had been in place for years. People were used to doing things a certain way, even when those processes were manual, fragmented, and creating unnecessary work.
What impressed me about Folio Tech wasn't just the automation. It was their ability to meet our team where they were. Instead of forcing us to rip and replace systems or completely change how people worked, they built around our existing processes and gradually introduced improvements that people could actually adopt.
The result was a significant reduction in manual work, better visibility into our operations, and a team that was finally spending time on higher-value activities instead of administrative tasks. It felt less like implementing software and more like adding experienced operators to the business.”