Blog / Industry Insights
William Charlesworth-Jones 9 min read Rework is the single largest hidden cost on a UK construction site. Most directors can quote the price of their last delivery of timber but couldn’t tell you, within a factor of two, what the site spent last quarter doing the same job twice. That’s not a personal failing. The industry has spent decades not measuring it.
Where the data does exist, it consistently lands in the same range: between 5 and 15 percent of total project value is lost to rework on a typical UK site. Autodesk’s Construction Disconnected research put the share of rework caused by poor project data and miscommunication at around half. The UK’s own Get It Right Initiative puts the total cost of error and rework, including indirect impacts, closer to 21 percent of project value when you count programme slippage and disputes alongside the direct redo cost.
On a £500,000 housing project, those numbers stop being abstract. They are the difference between a healthy year and a survivable one.
The figures vary by source and by project type, but the range across the credible UK and US construction research is consistent enough to plan from.
The implication for a director running an SME house builder or a small contractor is uncomfortable. If your projects average 18 percent margin and you lose 10 percent of project value to rework, then roughly half of every project’s profit is being eaten by work you have already paid for once.
Profit on paper is not profit in the bank. The work you build twice never shows up on an invoice. It shows up in the size of the year you have at the end of December.
If rework is that expensive, why isn’t it on the radar of every site manager and director the way that material cost or labour rates are? Because it doesn’t appear anywhere with the word “rework” stamped on it.
It hides in:
By the time you reconcile margin at the end of a project, the cause has been forgotten. The site team remembers the symptom (it was tight at the end) but not the source (a change made three months ago that never reached the trades on site).
Until you can see rework as a category in its own right, you can’t manage it. Until you can see where it came from, you can’t prevent it.
After dozens of conversations with site managers, directors, and trades across UK developments, the same five patterns come up over and over. They are not technical problems. They are coordination problems.
Drawing updates issued by email, attached to a WhatsApp message, or printed off and pinned in a cabin somewhere. By the time revision D is on the design coordinator’s screen, revision C is still being built off the rev C print on the brickie’s kitchen table from last Tuesday.
This is the single most common driver of physical rework. Walls in the wrong place. Apertures sized for the previous drawing set. Brickwork above DPC keyed in for a cladding type that has since changed.
A question raised on a Wednesday, answered verbally a week later, and the work has already moved on under an assumption the gang made on the spot. Sometimes the answer arrives, sometimes it doesn’t, and the assumption holds until it bumps into something on a later trade.
The cost lands when the next trade in shows up and finds the previous trade’s assumption was wrong.
The ironmonger’s order says brushed nickel. The site team heard “the silver one” on a walk-around. The silver one arrives. Doors hang. Owner walks the show flat. Doors come off.
Specifications travel badly when they travel verbally. Trades on UK sites are often working across two or three projects at once and have every reason to default to what they did on the last one.
An issue spotted on site at 11am. Photo taken on someone’s phone. Sent to a WhatsApp group, perhaps, then buried under the next forty messages. Three weeks later the issue is still there and now requires digging up.
Evidence not attached to the work item, not geo-located, not tied to a date and an author, has almost no value when the dispute arrives.
A CSCS card expired on Tuesday morning. The trade carrying it didn’t know. The site manager didn’t know. The audit caught it on Thursday afternoon. The work done in those three days now needs reviewing, possibly redoing, possibly defending against an HSE notice.
The cost here is rarely the inspection. It is the work it casts doubt on.
Take a moderately sized SME house builder running a £500,000 development. Margin 18 percent in a good year, so £90,000 of project profit before overheads if everything goes to plan.
Apply the midpoint of the rework range: 10 percent of project value lost to direct rework, which is £50,000. That is more than half of the project’s expected profit, gone to work that has been done twice.
Of that £50,000, around half, £24,500, traces back to miscommunication. Drawing revisions, RFIs, specs, photo evidence, certifications. The five sources above. Not technical complexity, not the inherent difficulty of the build. Information not reaching the right person at the right time.
To put £24,500 in a frame: that is roughly six to eight weeks of a skilled tradesperson’s fully loaded cost on the site. It is most of a year’s salary for a junior site administrator. It is more than a typical UK SME house builder spends on software, insurance, and marketing combined for a year.
And this is for one project. A director running three concurrent projects of similar size is, on the same logic, losing somewhere between £150,000 and £450,000 a year to rework, with half of that loss caused by information not flowing.
The miscommunication half of rework is the half that is achievable. You will not halve the technical complexity of UK groundworks. You can absolutely halve the rate at which the wrong revision of a drawing makes it onto a wall.
Five practical changes, ranked by impact and how quickly they can land:
When revision D is published, revision C should disappear from every screen at once. No email attachments to chase. No paper print-outs that have to be physically replaced. No WhatsApp messages with PDFs that get scrolled past.
This is the change that, on its own, attacks the largest single rework source we see. It also has the cleanest implementation: every member of site staff, including subcontractors, has the current drawing set on their phone, in their pocket, every day.
The cost of an RFI is not asking the question. It is the gap between the question and the recorded answer. Move the answer onto a written, shared, dated record and the gap closes.
This also has a knock-on benefit when a dispute arrives months later. The decision history is already on file.
Every issue raised should have a photo, attached to the specific task or item, with the location on site and a timestamp. Not in a group chat. Not on someone’s camera roll. On the record, with the work it relates to.
When the dispute about who pays for the redo arrives in week 14, the evidence is already where it needs to be.
“We’ll sort that next week” is the language of rework. “John to install the rev D bracket, Tuesday morning” is the language of work done once.
The shift from passive verbal commitments to a named owner, a specific date, and a written task closes the loop on a surprising fraction of the small issues that compound into big rework.
The CSCS card that expires on Tuesday should have been flagged the previous Monday. Treat compliance the way you treat MOTs: scheduled, reminded, renewed before the gap, not after.
This is the cheapest of the five fixes to implement and the one with the most direct downside if you don’t do it.
The math is straightforward and worth doing once on your own projects.
Take a director running three projects a year at around £500,000 each. Apply the rework range. Apply the miscommunication share. The figure that comes out is uncomfortably large: somewhere in the order of £70,000 to £75,000 per year of avoidable cost, attributable to information not flowing well enough across the site team.
Halve it (which is achievable with the practical fixes above), and the same director has somewhere between £35,000 and £37,500 of recovered margin to either keep, reinvest, or spend on something the business has been wanting to do for years.
That number is bigger than the cost of every piece of software the typical UK SME builder uses. Bigger than the cost of a part-time site administrator. Bigger, in most cases, than the next pay rise the director was hoping to give themselves.
The reason rework persists at industry-wide rates of 5 to 15 percent is not that anyone in UK construction is happy with it. It is that it is invisible, plural, and feels like the cost of doing business. None of those are reasons to leave it alone. They are reasons it is the highest-leverage cost a director can attack this quarter.
If you’d like to see how UK site teams are using BuildersAI to attack the miscommunication half of rework on their own sites, our demo is twenty minutes and includes a walked-through estimate of what you’d recover on your specific project size. Book a demo or have a look at how the app works on site.
Ready to fix this on your site?
BuildersAI puts drawings, tasks, and updates on every phone on site. 14-day free trial, every feature, no card.
William Charlesworth-Jones
Founder, BuildersAI
Founder of BuildersAI with a background in construction and economics. After watching coordination chaos waste hours on every site, he set out to build AI tools that actually work for UK builders.
LinkedInIndustry Insights
Poor communication is the single biggest hidden cost on UK construction sites. Here is what it is actually costing you, why it keeps happening, and what good looks like.
7 min readIndustry Insights
The UK construction industry needs nearly 48,000 new recruits every year. If you're already struggling to find decent trades, it's about to get harder. Here's what the numbers say and what you can do about it.
10 min readGet started
14-day free trial with every feature, no card and no contract, and your team can be set up within a day.
