Your ERP does a lot. It manages commitments, tracks costs, runs payroll, handles AP/AR, and keeps your project financials in one place. That's not nothing. For a general contractor running 15 active projects, a good ERP is infrastructure.
But pre-construction isn't execution. And the tools built for execution have a hard time shifting backwards in the timeline to where decisions still matter.
Pre-con is where you're reading 2,000-page spec books, cross-referencing drawings against sub quotes, and trying to write scope packages tight enough that nobody comes back to you after award. Your ERP has nothing to say about any of that.
That's not a knock on ERPs. It's just not what they were built for. The problem is that this gap — between what your ERP covers and what pre-con actually requires — is where most of your margin risk lives.
To be fair, most modern ERPs have made some moves into pre-construction. Viewpoint, CMiC, Sage, and others offer bid management modules, budget templates, and subcontractor databases. Some have estimating integrations. A few have added RFQ workflows.
These are useful. But they're workflow containers — not analytical tools. Here's what that means in practice:
The left column is data management. The right column is document intelligence. Your ERP does the first. It cannot do the second.
The ERP-to-field handoff is well-understood. But the pre-con gap — the space between "we're bidding this" and "we have a signed subcontract with complete scope" — is where GCs consistently bleed margin.
According to the Arcadis 2025 Global Construction Disputes Report, the average U.S. construction dispute value hit $60.1M in 2024. For six of the last nine years, the top cause has been errors and omissions in contract documents. That's not a field problem. That's a pre-con problem.
The numbers at the project level are just as concrete. Scope gaps drive 8–14% in change orders on typical commercial work, according to Navigant research republished by the AIA. On projects with weak scope packages, that number climbs past 25%.
Provision's own research across 200+ GC interviews backs this up with specific examples that most estimators will recognize immediately:
None of those gaps would have shown up in an ERP report. They all lived in the documents — the drawings, the specs, the supplementary conditions — that the ERP never read.
As one Pre-Construction Lead at a Top-ENR Canadian GC put it: "If you miss anything, they'll bill it."
Here's the core issue. Pre-construction is a document problem. A typical commercial bid package includes:
A complete project set can run 1,500 to 2,500 pages. Estimators read it all — or they're supposed to. In practice, under bid-day time pressure, things get missed. That's not a failure of discipline. It's a volume problem.
ERPs don't read documents. They store data that humans enter after the documents have been reviewed. The review itself — the hardest, highest-stakes part of pre-con — has always been entirely manual.
That's exactly where purpose-built pre-con AI is designed to work.
Provision's Chat Agent reads the complete project set — drawings, specs, contracts, RFIs, and addenda — and answers estimator questions with cited responses in under 20 seconds. Not "check Division 7." The exact clause, the exact page, the exact drawing reference.
That's not something your ERP can do. It's also not something a general-purpose AI like ChatGPT does reliably — generic AI lacks the construction context to handle project-set ingestion and produce the structured, cited outputs estimators depend on.
Provision has answered 50,000+ estimator queries across live bid projects. That's 50,000 questions that would have required someone to manually search a spec book or flip through drawing sets.
Your ERP has subcontract templates. That's a starting point. But a good scope-of-work package isn't a template — it's a document-specific extraction of every requirement that applies to a given trade on this project.
Provision's Scope Agent generates complete scope-of-work packages from construction documents in under 60 minutes. It replaces 30–40 hours of manual work per bid. It achieves 97% match accuracy on scope item extraction — compared to a human estimator baseline of 91.3% on the same exercise.
The Scope Gap Playbook — based on 200+ GC interviews — identifies the most common anti-pattern in scope writing: "As per plans and specs." That phrase is how GCs lose money. It's a scope gap dressed up as a scope item. Scope Agent doesn't produce that. It produces specific, document-referenced line items by trade.
Most ERP contract modules store templates and track execution status. They don't read a specific owner contract and flag the clause on page 47 that shifts site safety responsibility to the GC in an unusual way.
Provision's Risk Review does. It reviews contracts and specifications for risk with 99% accuracy on pre-built checklists, cuts review time by 80%, and cites every flagged risk to the exact clause. For GCs reviewing owner agreements with supplementary conditions, that's the difference between knowing your exposure before you sign and discovering it during a dispute.
The Arcadis data is worth repeating here: $60.1M average dispute value. "Errors and omissions in contract documents" as the leading cause for six of nine years. That's the risk Risk Review is built to catch — before it becomes a dispute.
One of the most consistent findings in Provision's GC research was the handoff problem. As a Director of Pre-Construction at a mid-market Southeast GC described it: "Pre-con is working in the scope sheet world and project management is working in the scopes of work."
ERPs can't bridge that gap because they don't understand the documents the scope came from. When a PM needs to know why a particular item is or isn't in a sub's scope, they're usually going back to the same documents the estimator reviewed months earlier — and the ERP has no record of that analysis.
Purpose-built pre-con AI creates a documented, searchable trail. What was reviewed. What was flagged. What was included and why. That's not a workflow enhancement — it's a risk reduction tool for every project that goes into execution.
To be direct: Provision doesn't replace your ERP. It fills the gap your ERP was never built to fill.
Your ERP handles commitments, costs, and compliance once the project is awarded. Provision works on the front end — bid review, scope generation, risk identification, document Q&A — so that what flows into your ERP at award is already clean.
Think of it this way. Your ERP manages the contract. Provision helps you write a contract worth managing.
EllisDon — one of Canada's largest GCs — used Provision to save $1.8M in a single pre-construction cycle. That number came from scope gaps caught before buyout, not from execution efficiencies. Read the EllisDon case study for the full breakdown.
The Scope Gap Playbook identifies eight habits that separate high-margin GCs from average ones in pre-construction. Several of them speak directly to what AI tools can now support:
For a full breakdown of these habits — and the anti-patterns that drive scope gaps — see the Scope Gap Playbook.
ERP vendors are adding AI features. That's real. But the AI being added to execution-layer tools is primarily about automating cost reporting, change order approval workflows, and financial forecasting. It's not reading drawings. It's not generating scope packages. It's not catching the $300K lead-lined glass omission buried in Division 8.
Pre-construction is a document-intelligence problem. The tools being built to solve it are purpose-built for that specific workflow — not adapted from accounting software.
Provision has reviewed more than $100 billion in project value across 100,000+ documents. That's not a proof of concept. It's a track record.
If your ERP is your only pre-con tool in 2026, you're still doing the hardest part manually. See what fills the gap — or explore how Provision works for general contractors.
The preconstruction gap is the space between what ERPs do — manage project data, costs, and commitments — and what pre-con actually requires: reading construction documents, identifying scope gaps, generating scope packages, and flagging contract risk. ERPs were built for execution, not document-level pre-con analysis.
No. ERPs store data that humans enter after reviewing documents. They do not ingest or interpret drawings, specs, or contracts. That document-intelligence function — reading a full project set and surfacing what matters by trade — requires purpose-built tools designed for that specific workflow.
Purpose-built pre-con AI reads the full project set, generates scope-of-work packages, answers estimator questions with cited references, and flags contract risk tied to specific clauses. Provision's products do all of these. They work alongside your ERP — not instead of it — by cleaning up what enters the system at award.
Scope Agent generates complete, document-specific scope-of-work packages from construction drawings and specs in under 60 minutes. It replaces 30–40 hours of manual work per bid and achieves 97% match accuracy on scope item extraction. No ERP module does any of that — they provide templates, not extraction from live bid documents.
Scope gaps drive 8–14% in change orders on typical commercial projects, according to Navigant/AIA research. On projects with weak scope packages, that number exceeds 25%. The Arcadis 2025 Global Construction Disputes Report puts the average U.S. dispute value at $60.1M, with errors and omissions in documents as the leading cause for six of the last nine years.
No. Provision works in pre-construction — bid review, scope generation, document Q&A, and contract risk identification. Your ERP manages project execution after award. Provision helps ensure what flows into your ERP at award is already complete and accurate. The two tools serve different parts of the project lifecycle.
Generic AI lacks the construction context required for project-set ingestion, scope extraction, and structured bid outputs. It doesn't understand drawing hierarchies, trade-specific scope requirements, or how specs interact with supplementary conditions. Purpose-built tools are trained on construction workflows and produce the cited, structured outputs that estimators actually need on bid day.
Provision reads your full project set and catches scope gaps before bid day.
Book a demoMore Articles
