Rework is everywhere on construction projects, and almost everyone underestimates it. The visible costs \u2014 demo, re-installation, material waste \u2014 make it onto a cost report. The invisible costs \u2014 superintendent time, crew standby, schedule compression, subcontractor premium for out-of-sequence work \u2014 mostly don’t. Add them together and rework is consistently one of the largest single line items on projects that seem to be tracking fine on paper.<\/p>\n
The harder question is where it comes from. Not all rework is the same. Some is genuinely unavoidable: owner scope changes, unforeseen conditions, genuine design evolution. But a large share traces directly to information problems \u2014 drawings that don’t agree with each other, specs that contradict the details, coordination gaps between disciplines that no one reconciled before the trades hit the field.<\/p>\n
That share is preventable. This post explains how to size the problem, where rework actually originates, how it compounds into change orders and RFIs, and where in the project timeline the fix is cheapest.<\/p>\n
Rework is notoriously hard to measure precisely \u2014 most of it gets absorbed into productivity losses and labor overruns that never appear as a discrete line. Industry studies estimate it runs in the range of 5\u20139% of contract value on typical commercial projects, though project teams almost uniformly believe their figure is lower than the actual. Some of the gap is accounting: when a crew re-hangs ductwork because the structural framing wasn’t where the drawing said it would be, the cost usually lands in labor variance, not a rework line.<\/p>\n
What’s more useful than an aggregate percentage is understanding the cost curve. A conflict caught during plan review \u2014 in the documents, before anything is built \u2014 costs the time it takes to issue a coordination note and a revised sheet. The same conflict caught in the field, after the work is installed, costs demolition, re-work, a change order to cover the extra labor and material, and schedule time while the crew works around the problem.<\/p>\n
Flikt’s own data, documented across real projects, puts that ratio at roughly ten to one<\/strong>: an error that costs one unit of effort to fix in review costs ten units to fix in construction. That’s not a marketing claim \u2014 it’s the basic economics of concrete versus paper.<\/p>\n Construction rework has multiple root causes<\/a>, and the right prevention strategy depends on which category you’re dealing with.<\/p>\n When an owner changes program mid-design or adds scope after construction starts, rework is the price. There’s limited mitigation here \u2014 these are deliberate decisions. The goal is to push them as early as possible, and to price them at design time rather than construction time.<\/p>\n Existing-conditions surprises \u2014 buried utilities, soil bearing failures, structural conditions hidden in walls \u2014 generate rework that no amount of document review prevents. Pre-construction investigation reduces it; it never eliminates it.<\/p>\n This is the preventable category, and it’s larger than most project teams account for. It covers:<\/p>\n On one real project in Flikt’s data, more than 40 change orders<\/strong> traced directly to conflicts between<\/em> 2D plan sheets \u2014 not scope changes, not site surprises, not bad weather. Documents that didn’t agree with each other. One structural-MEP conflict alone \u2014 a shear wall colliding with duct routing \u2014 carried $28,000\u2013$45,000 in rework plus 14\u201321 days of schedule<\/strong>.<\/p>\n That single conflict, caught during plan review, is a coordination note. Caught in the field, it’s a six-figure line item by the time you add general conditions and schedule compression.<\/p>\n These three aren’t the same thing, but they’re tightly connected, and the total cost of a coordination failure is usually all three together.<\/p>\n An undetected conflict typically surfaces first as an RFI<\/strong> \u2014 the field asks for clarification. If the answer to the RFI requires physical changes to work already installed, it becomes rework<\/strong>. If the rework costs more than the original scope, it becomes a change order<\/strong>. Each step in that chain adds cost and schedule: the RFI eats superintendent and PM time; the rework eats trade labor and material; the change order triggers a pricing and approval cycle before work can restart.<\/p>\n The true cost of change orders in construction<\/a> breaks out the administrative overhead \u2014 the project management hours, schedule float consumed, and subcontractor premium for out-of-sequence work \u2014 that rarely appears in the change order value itself. On complex projects, the indirect cost can exceed the direct cost.<\/p>\n The thread through all of it: a single coordination failure in the documents typically generates an RFI and<\/em> a change order and<\/em> rework, in sequence. Preventing the document error prevents all three. That’s why the cost of change orders<\/a> and the cost of rework are so closely correlated on the same projects.<\/p>\n The cost of fixing a conflict is lowest before anything is built, and it drops sharply the earlier you catch it. The progression looks like this:<\/p>\n In design:<\/strong> A conflict surfaced before documents are issued is a coordination note. The designer updates the sheet; no field impact.<\/p>\n At bid \/ design review:<\/strong> A conflict surfaced before mobilization costs a clarification and a revised issue. No demo, no out-of-sequence crew work, no field disruption.<\/p>\n During construction, work not yet installed:<\/strong> Typically an RFI plus minor replanning. Some schedule impact; no demolition.<\/p>\n After work is installed:<\/strong> Demo, reinstall, change order, schedule impact, potential follow-on trade impacts. Ten-to-one ratio applies here.<\/p>\n This is why pre-construction QA<\/a> is the highest-leverage intervention. The same effort that catches twenty coordination gaps during plan review would catch \u2014 at best \u2014 twenty change orders during construction, each requiring a separate resolution cycle, field disruption, and cost. The work is identical; the unit cost is not.<\/p>\n The practical implication: a thorough document review immediately before bid, while the full set is issued but before mobilization, is the window where the ten-to-one ratio works in your favor. Anything caught there costs a fraction of the same catch in the field.<\/p>\n The reason coordination failures persist isn’t that project teams don’t know they’re expensive. It’s that finding them in a large drawing set is slow, manual, and fatigue-prone. A senior reviewer reading hundreds of sheets against each other \u2014 structure against MEP, MEP against architecture, drawings against specs \u2014 will catch most of the high-probability conflicts and miss some of the cross-sheet ones that require holding two details in mind simultaneously.<\/p>\n Flikt runs AI construction plan review<\/a> directly on 2D PDF drawing sets \u2014 no BIM, no model exports, no modeling staff required. It reads the full document set the way a comprehensive review should: every sheet against every other sheet, drawings against specs, and across all disciplines simultaneously. It surfaces the spec-vs-drawing mismatches, cross-sheet conflicts, cross-discipline spatial conflicts, and missing-coordination gaps described above.<\/p>\n The benchmark Flikt uses to characterize drawing quality is conflicts per 100 sheets<\/strong> \u2014 the count of real, actionable coordination issues normalized by drawing-set size. It’s a concrete, comparable metric that lets owners and GCs evaluate drawing quality across design teams and project revisions. The data behind it \u2014 real conflict counts from real projects, including the 40-change-order case above \u2014 is on the evidence page<\/a>.<\/p>\n For a full breakdown of conflict taxonomy and what the review catches versus what BIM clash detection catches, see clash detection without BIM<\/a> and the conflicts per 100 sheets benchmark<\/a>.<\/p>\n If you’re a general contractor seeing chronic rework on projects, the diagnostic starts with source attribution: what share of your change orders and rework traces to coordination failures in the documents, versus scope changes, versus unforeseen conditions? If it’s more than a third coordination-related, the intervention point is plan review \u2014 not field process, not subcontractor selection.<\/p>\n The right sequence:<\/p>\n For GCs who want to take this into a formal QA process, the pre-construction QA checklist<\/a> and constructability review<\/a> cover the full workflow.<\/p>\n If you’re an owner or developer who’s absorbing rework costs through GC contingency and schedule, the conversation to have is with your design team \u2014 and the question is how comprehensively the documents were reviewed before they were issued. The cost of rework in construction is largely a cost of information problems, and information problems are caught in review.<\/p>\n Flikt reviews your 2D plan sets for cross-discipline conflicts before construction \u2014 no BIM required, no modeling staff. Start with AI construction plan review<\/a>, review the documented results on the evidence page<\/a>, or run the numbers for your own project<\/a>.<\/em><\/p>\n Industry figures cited in this article are traced to their primary documents in our construction rework statistics register<\/a>, which also lists the widely-repeated numbers that did not survive checking.<\/p>\n Industry studies estimate rework runs roughly 5 to 9 percent of contract value on typical commercial projects, and project teams almost uniformly regard that as an underestimate, because much of it is absorbed into productivity losses and labor overruns that never appear as a discrete line item.<\/p>\n Substantially cheaper, though no reliable published multiplier exists \u2014 the widely-quoted ten-times rule is not a construction measurement. What is measurable is that design errors and omissions account for 1.5% of contract value in field rework on building projects, and that a conflict resolved on paper costs a markup and a coordination call, while the same conflict in the field carries demolition, rework, a change order and schedule impact. Construction, once demolition, reinstallation, change-order administration, and schedule impact are counted.<\/p>\n Information problems in the documents. Four categories account for most of it: spec-vs-drawing conflicts, where the drawing details one product and the specification calls another; cross-sheet conflicts, where dimensions or elevations do not reconcile between sheets; cross-discipline conflicts, where the seams between structure, mechanical, electrical, and architecture do not align; and missing-coordination gaps, where something required is simply absent from the sheet it should appear on.<\/p>\n They compound in sequence. An undetected conflict usually surfaces first as an RFI when the field asks for clarification. If the answer requires physical changes to work already installed, it becomes rework. If that rework costs more than the original scope, it becomes a change order. A single coordination failure in the documents typically generates all three, so preventing the document error prevents all three.<\/p>\n Before anything is built, and the cost drops sharply the earlier it is caught. In design it is a coordination note with no field impact. At bid or design review it costs a clarification and a revised issue. During construction, on work not yet installed, it is typically an RFI plus minor replanning. After work is installed it means demolition, reinstallation, a change order, and schedule impact, which is where the ten-to-one ratio applies.<\/p>\n Yes. Flikt.AI runs AI construction plan review directly on 2D PDF drawing sets, with no BIM, no model exports, and no modeling staff required. It reads every sheet against every other sheet and drawings against specifications, across all disciplines, and reports conflicts per 100 sheets as a normalized drawing-quality benchmark.<\/p>\n Upload your plan set and get AI-powered conflict detection \u2014 no BIM required.<\/p>\n Get Started →<\/a> Industry rework figures describe the average project, which is nobody’s project. If you would rather measure your own, send a completed set and the RFI log it produced<\/a>.<\/p>\n
\nWhere rework actually comes from<\/h2>\n
Scope changes and owner decisions<\/h3>\n
Unforeseen site conditions<\/h3>\n
Design errors and coordination failures<\/h3>\n
\n
\nRework, change orders, and RFIs: how they compound<\/h2>\n
\nWhere the fix is cheapest: the pre-construction window<\/h2>\n
\nFlikt’s approach: find the conflicts in the documents<\/h2>\n
\nWhere to focus if rework is already a problem on your projects<\/h2>\n
\n
\nFrequently asked questions<\/h2>\n
How much does rework cost on a construction project?<\/h3>\n
How much cheaper is it to catch a conflict in plan review than in the field?<\/h3>\n
What causes preventable rework in construction?<\/h3>\n
How are rework, change orders, and RFIs related?<\/h3>\n
When is the cheapest point to fix a coordination conflict?<\/h3>\n
Can you find conflicts between construction documents without a BIM model?<\/h3>\n
Ready to catch conflicts early?<\/h3>\n
\n📅 Book a Demo<\/a>\n<\/div>\n