Revision C Was Issued. The Field Was Still on B.
Quick answer: Much of the cost of rework traces back to information. In a 2018 FMI and PlanGrid survey, respondents attributed 48% of US rework to poor project data and miscommunication, about $31.3 billion that year. A 2026 study of one contractor's projects found that actual rework averaged 0.76% of contract value once post-completion corrections were counted, with one project at 7.34%, and that the contractor's own records had under-reported pre-completion rework by 300%. Marking one revision as current, logging who received it, and flagging mismatches before the pour gives a crew a record to check before the work gets done twice.
Key Takeaways
- ●Survey respondents attributed 48% of US rework to poor project data and miscommunication, about $31.3 billion in 2018 (FMI and PlanGrid).
- ●In one contractor's project data, actual rework averaged 0.76% of contract value including post-completion corrections, and 7.34% on its worst project (Love, 2026).
- ●The same study found the contractor's own records under-reported pre-completion rework by 300%, so much of it never gets recorded as rework.
- ●Autodesk and FMI estimated in 2021 that decisions made on bad data may account for 14% of global rework.
- ●In one ServiceIQ implementation, a new revision is marked current, records tied to the prior revision are flagged, transmittals serve as time-stamped proof of delivery, checklists cite the revision they were completed against, and the morning briefing lists revisions a sub hasn't acknowledged.
What the Numbers Say, and How Far to Trust Them
Three studies cover this, and each has limits. FMI and PlanGrid surveyed nearly 600 US construction leaders in 2018, and the 48% and $31.3 billion are what respondents attributed to poor project data and miscommunication. That's dated and survey-based.
Peter Love's study in the Journal of Construction Engineering and Management (January 2026) looked at one contractor's projects, so it describes that contractor and doesn't speak for the industry. Its quality managers set out to measure what the company's books had missed. Pre-completion rework averaged 0.38% of contract value, and adding post-completion corrections brought it to 0.76%, with a range from 0.02% to 7.34% across projects.
Autodesk and FMI's 2021 report, based on more than 3,900 construction professionals, estimated that decisions made on bad data may account for 14% of global rework in 2020. That one is a modeled estimate.
The studies use different methods, regions, and years. They land in the same place, which is that information problems sit behind a large share of the rework contractors pay for.
Why Rework Is Hard to See
The finding in Love's study that matters most to an owner is the under-reporting. The contractor's own records had missed pre-completion rework by 300%, so most rework never gets written down as rework. It shows up as a crew back on a wall, a change order nobody connected to it, or a margin that came in short.
At many firms the knowledge of who holds which revision lives informally with the superintendent. When that person moves on, the replacement doesn't inherit it.
What Changes When the Current Revision Is the Only One Anyone Can Find
In one ServiceIQ implementation (September 2026), a new revision is marked current and every record still tied to the prior one gets flagged. Transmittals go out as proof of delivery, so "we never got Rev C" has a timestamp. Checklists cite the revision they were completed against, which makes a mismatch visible before the pour. The morning briefing lists revisions that were issued without an acknowledgment from the sub.
Document control, RFIs, and submittals are included on every ServiceIQ plan.
What It Doesn't Do
It works from what's in the system. A revision that never gets uploaded is invisible to it. It shows which sub hasn't acknowledged a revision, and it can't make anyone read it. It also doesn't change what a crew builds from if they never open the current drawing.
Where to Start
Pick one problem and baseline it before you buy anything. Go back through your last few jobs and count the rework incidents that trace to a revision or a spec change. Then time how long it takes from issuing a revision to each sub acknowledging it. Love's own recommendation is to track rework costs and their causes, because the records are where the gap shows up.
Then ask any vendor three questions. Which of my systems does it read? What does it do when it finds something? Who approves?
Revision changes often arrive alongside change orders, which "The Fee You Lose Between the Field Text and the Change Order" covers. "The Complete Guide to Choosing Field Service Software" has the wider vendor checklist.
The slides from our EdTech Summit session, "AI in the Construction Workplace," cover this problem and four others, and they're a free download.
Frequently asked questions
Keep reading
The Fee You Lose to Change Orders
98% of GCs have lost fee to change-order negotiations, mostly in the gap between a field change and the paperwork catching up. Here's where it goes.
Closing the Punch List Gap
A GC losing $18,000 to a missing photo and a foggy memory. Owners price a punch list holdback around how long it takes to get proof, so here's what changes when proof is instant.
AI in the Construction Workplace: Five Problems
Five problems come up on nearly every contractor's desk, change orders, information overload, rework, slow payment, and knowledge loss, and the AI that fits each.
Ready to take on more premium projects without adding to your overhead?
Lock in your flat-rate, unlimited-user fee with ServiceIQ today.
Start your free trial