Cost of Rework in Construction: 4 to 12% of Project Value

Updated June 2026. Sources: CII IR-153, NIST GCR 04-867, GIRI UK research (2015/16).

The canonical datasets

Three pieces of research carry most of the citation weight when construction rework cost comes up. The Construction Industry Institute (CII), based at the University of Texas at Austin, published Implementation Resource 153 (IR-153), An Investigation of Field Rework in Industrial Construction, summarising field rework data from a database of 144 industrial construction projects. IR-153 remains the most-cited construction rework benchmark in North America.

The National Institute of Standards and Technology General Construction Report GCR 04-867 (2004), Cost Analysis of Inadequate Interoperability in the US Capital Facilities Industry, estimated $15.8 billion in annual losses to the US construction industry from interoperability failures, the majority of which manifested as rework. That interoperability estimate remains the most-cited dollar figure for US construction waste of this kind.

The Get It Right Initiative (GIRI) is a UK-based industry consortium whose foundational research (carried out in 2015/16, the basis for the initiative launched in 2017) calculated that UK construction loses 10 to 25% of project cost to avoidable error, averaging around 21% once indirect costs and latent defects are counted, with the direct avoidable-error cost alone at roughly 5% of project value. GIRI's total figures are higher than the US CII direct averages, partly because GIRI counts the cost of defective work left in place (which CII excludes) and partly because the UK study captured residential as well as industrial work.

What the named datasets report

Note: the direct figures count field rework only and exclude design rework (typically 1 to 3 percentage points more) and the cost of defective work left in place. Delivery model matters: modular and prefabricated construction consistently reports lower field rework than traditional design-bid-build, because factory conditions remove much of the on-site variability, though no single public percentage is authoritative.

The five causes (CII categorisation)

The CII field-rework studies consistently rank the same five primary causes. In rough order of total cost share:

1. Owner or client changes (roughly 33% of rework cost). The single largest cause across most project types. Late changes to scope, specification, or finish drive cascading rework across trades. Front-end planning quality is the strongest predictor of how much owner-change rework a project will absorb.
2. Design errors and omissions (roughly 28%). Drawings that conflict, specifications that miss critical detail, dimensions that do not coordinate across disciplines. BIM clash detection has reduced this category meaningfully on projects that use it well, but it remains a major rework driver on traditional CAD-based projects.
3. Communication failures between trades (roughly 17%). The mechanical contractor and the electrical contractor each assume the other will route around a shared space; both build, both have to redo. Disciplined coordination meetings and shared 3D models reduce but rarely eliminate this category.
4. Defective materials and equipment (roughly 12%). Out-of-spec deliveries, damaged stock installed before inspection, equipment failures during commissioning. The supply-chain volatility of 2022 to 2024 pushed this category up on many projects.
5. Worker error (roughly 10%). Often traceable upstream to training gaps or supervision gaps rather than worker fault. The most actionable interventions are toolbox talks tied to specific recurring error categories and supervisor sign-off discipline on critical-path activities.

The prevention return

The Construction Industry Institute Best Practice for Front-End Planning (BP-1989) reports total project cost reductions of 5 to 15% from disciplined front-end planning, with the rework reduction component being the largest single source of those savings. Within that range, the projects with the highest front-end planning scores (as measured by the Project Definition Rating Index, PDRI) consistently show rework cost at the low end of the CII rework distribution, often below 3% of project value.

The directional finding from GIRI's UK research is similar: the prevention dollar pays back roughly 5 to 10 times in avoided rework, which is consistent with the Boehm cost-of-change curve in software (where the prevention-to-failure ratio is also reported as 1:10 to 1:100 depending on the development phase). This cross-industry consistency is one of the more interesting findings in the cost-of-quality literature: whatever the medium being built, the ratio of cost-to-prevent versus cost-to-rework lands in a remarkably narrow band.

For owners, the practical takeaway is that the highest-leverage spend on a major capital project is the front-end definition phase, when costs are still measured in tens of thousands of dollars rather than tens of millions. The CII PDRI and the FHWA front-end planning frameworks both provide structured ways to spend that money well. For contractors, the highest-leverage internal investment is usually in the quality system that catches drawing conflicts and trade coordination errors before they go to field.

How construction rework compares to software rework

The headline construction rework rate (4 to 12% of project value) is meaningfully lower than the headline software rework rate (20 to 40% of total development effort per Capers Jones, Applied Software Measurement). The gap is real, but it reflects two structural differences rather than a difference in execution quality.

First, construction defects are physically obvious. A wall in the wrong place gets noticed by an inspector; software defects can hide in code paths nobody exercises. Construction has lower latent defect rates partly because hiding defects is harder. Second, the cost of late changes in construction is so much higher than in software that owners self-select into doing more front-end planning. The 5% construction rework rate exists because the alternative (a 30% rework rate on a $100M building) would bankrupt the owner. Software teams routinely run the rework rate that would be unaffordable in construction, because the medium absorbs it.

The cross-industry lesson runs both ways. Software teams have something to learn from construction's front-end planning discipline (see the unclear-requirements page). Construction has something to learn from software's iterative review cycles, particularly for early scope discovery on owner-led projects. The software rework page and the manufacturing rework page sit alongside this one for cross-industry reference.

Sources

• Construction Industry Institute. Implementation Resource 153: An Investigation of Field Rework in Industrial Construction (144-project database). CII, University of Texas at Austin.
• Love, P.E.D. and Li, H. Quantifying the Causes and Costs of Rework in Construction. Construction Management and Economics, 18 (2000), 479-490.
• NIST. GCR 04-867: Cost Analysis of Inadequate Interoperability in the US Capital Facilities Industry. NIST, 2004.
• Get It Right Initiative. Research Report. GIRI, 2015/16 (initiative launched 2017). getitright.uk.com.
• Construction Industry Institute. Best Practice 1989: Front-End Planning. CII, University of Texas at Austin.
• Modular Building Institute industry commentary on lower rework in modular and prefabricated construction.

Frequently asked questions