What a buildability review actually changes on architect's drawings: 7 real examples from RIBA Stage 3-4
Our refurbishment of Grade II listed Georgian former private bank in Belgravia
Quick answer: A buildability review is a contractor-led, structured read of your design, usually at RIBA Stage 3 or early Stage 4, that flags where the drawings will cost more, take longer, or quietly fail on site than the design team realises. Done properly, it typically produces 20-40 specific comments per package: most minor, a handful worth six-figure sums. Below are seven real before/after examples from projects we've worked on, showing exactly what got changed and why.
What is a buildability review, really?
Most definitions you'll find online describe buildability as "the extent to which a design facilitates construction." That's accurate and useless.
In practice, a buildability review is a contractor sitting down with your tender-ready drawings and specifications, and producing a list of specific changes that, if made now, will save you money, time, or a warranty argument three years after handover.
It isn't a cost plan. It isn't a risk register. It isn't value engineering, though it shares DNA with all three. It's closer to a code review: someone who will actually have to build the thing telling you where your drawings will meet reality badly.
When should it happen?
The sweet spot is late RIBA Stage 3 through early Stage 4:
At Stage 3 the design is concrete enough to review but changes are still cheap. By Stage 4 you're issuing for tender. By Stage 5 you're arguing about who pays for the change.
If you only do one review, do it at the Stage 3/4 boundary. If you can afford two, do a light one at Stage 2 to flag principles, and a detailed one at 3/4 to work through packages.
The 7 examples
1. The £18,000 steel beam
Before: A double-height retail unit needed a column-free 9m span at ground floor. The structural engineer specified a cranked 686×254 UB with a bolted moment connection, allowing the cladding line to step forward at the transfer.
After: The cranked fabrication required a specialist sub-contractor with an 11-week lead time, and the moment connection needed site welding that would have pushed steel erection onto a second crane visit. A slightly deeper 762×267 straight UB with a simpler bearing detail, and the cladding line adjusted inward by 40mm, saved £18,000 in fabrication and three weeks in programme.
Why this keeps happening: Structural engineers optimise for load paths. Architects optimise for visual lines. Neither is pricing fabricator lead time or crane hours.
2. The basement tanking that would have leaked
Before: A single-skin Type A bitumen-based tanking system for a basement apartment, with construction joints at standard 5m intervals.
After: The site investigation showed a perched water table 1.2m below ground level with seasonal variation. Type A alone relies on perfect workmanship at every construction joint, achievable in theory, risky in practice for a client with a 12-year warranty. We recommended a Type C cavity drain membrane as a backup system to BS 8102 Grade 3. Extra cost: £42/m² over 180m². The alternative was a latent defects claim.
Why this keeps happening: Waterproofing specifications often default to the cheapest compliant option. Building Regs don't explicitly require belt-and-braces; an insurer usually does.
3. The curtain walling reveal that wouldn't close
Before: Elevations showed a 10mm shadow gap between the curtain walling mullion and adjacent SFS framing, with a flush plaster return.
After: Structural frame deflection under dead load + installation tolerance + plaster build-up meant the realistic achievable gap was 25mm minimum. We produced a revised detail with a 30mm reveal and a pre-formed metal trim. Caught at Stage 4, this is a drawing update. Caught on site, it's a snagging argument that drags into handover.
Why this keeps happening: Design details often assume zero tolerances. Buildings move. Frames deflect. The gap you drew is never the gap you get.
4. The MEP riser that didn't fit the services
Before: A residential MEP riser drawn at 600mm × 400mm, serving six floors of a 24-unit scheme. SVP, rainwater, sprinklers, boosted cold water, two electrical risers, comms and a smoke extract all routed through it on coordination drawings.
After: When we ran the services through the MEP sub's 3D model at actual outside diameters with required access clearances, the minimum workable riser was 900mm × 600mm. The architect revised the plan — losing ~0.4m² from each apartment on that stack but keeping the scheme buildable. The alternative was splitting the riser, which meant two shaft penetrations per floor plate and a structural redesign.
Why this keeps happening: Riser sizes are often set early, by eye, from previous projects. Modern residential carries more services than it did five years ago, particularly with heat pumps and MVHR now standard.
5. The glazed facade with nowhere to stand a scaffold
Before: A four-storey mews redevelopment with glazed facades to three elevations, each within 400mm of the site boundary.
After: There was no room to erect a conventional scaffold without licences from three separate neighbours, one of them a school with term-time restrictions. We costed a mast climber at £35,000 plus £6,000/month hire, versus phasing the glazing install to follow the structural sequence with a hoist. The phased option added four weeks to programme but saved £28,000. The decision was needed at Stage 4 because it affected procurement of the glazing package.
Why this keeps happening: Access strategy is treated as a contractor problem. It isn't, it's a design constraint that should shape the sequence and sometimes the detail.
6. The brick coursing that would have wasted 400 cuts
Before: Window head heights set at 2,150mm and 2,310mm across the main elevation. Standard 215mm brick coursing plus 10mm perp joint = 75mm module.
After: Neither head height hits a course. At 2,150mm you're 10mm off; at 2,310mm you're 15mm off. Across 28 windows that's roughly 400 cut bricks — each one extra labour, extra waste, and a visible half-brick somewhere in the elevation. A 5mm adjustment at each head (2,145mm and 2,325mm) resolved the coursing and lost nothing visually. Sills, DPC and parapet were checked at the same time and coursed correctly.
Why this keeps happening: Window heights get set to regulatory minimums and round numbers. Brick coursing is rarely checked at GA stage.
7. The rooflight upstand that voided the warranty
Before: A single-ply membrane roof with a pyramid rooflight; detail showed the membrane turned up the upstand by 100mm.
After: The membrane manufacturer's warranty (25 years) required a minimum 150mm upstand above finished roof level. The shown detail would have been installed as drawn, passed visual inspection, and failed the warranty audit on completion. The upstand had to grow from 200mm to 250mm, which meant the rooflight callout height had to increase — a change to the rooflight supplier's order. Caught at tender, this is a specification edit. Caught during installation, it's a delivery delay and a change notice.
Why this keeps happening: Manufacturer-specific warranty requirements vary by product and change over time. Architects hold them in spec documents; they don't always migrate into detail drawings.
What to bring to a buildability review
To get the most out of a contractor's buildability review, send over in advance:
GA plans, sections and elevations at a scale the details can be read at
Key detail drawings, especially the junctions: ground, eaves, parapet, rooflight, window head/cill/jamb, wet rooms, thresholds
The structural engineer's scheme drawings and loading assumptions
MEP coordination drawings, or ideally the federated model
Specifications for external envelope, waterproofing, and any warranted systems
Site constraints, access, boundaries, working hours, neighbour conditions
Programme assumptions
A good reviewer will come back with a marked-up set and a structured comment log - typically 20 to 40 items, categorised by cost impact, programme impact, and risk. Most will be minor. A few will be the difference between a clean handover and a two-year dispute.
FAQ
What's the difference between a buildability review and value engineering? Value engineering focuses on cost reduction; buildability focuses on constructability and risk. A buildability review will often surface VE opportunities, but its primary purpose is to find things that will cause problems on site, some of which cost more to fix properly, not less.
When is the latest I can do a buildability review? Practically, Stage 4 before tender. After that, changes trigger re-pricing, re-drawing, and sometimes re-approval. Stage 5 reviews are possible but the ratio of "things we can fix cheaply" drops fast.
Do I need a main contractor to do a buildability review? Not necessarily. An experienced project manager or independent buildability consultant can do it, but a contractor who actually prices and builds work in your sector will find more, faster. Early Contractor Involvement (ECI) models are designed around exactly this.
How long does a buildability review take? For a £2–5m mid-rise scheme, expect 3–5 days of a senior person's time plus a review meeting. For a £10m+ commercial or mixed-use scheme, 2–3 weeks including sub-contractor input on specialist packages.
How much does a buildability review cost? Under a PCSA, typically £8,000–£25,000 depending on project size and scope. If provided informally by a contractor bidding for the job, often free — but expect a lighter review.
Does a buildability review replace a contractor's tender queries? No. It's upstream of that. A good buildability review reduces tender queries — and therefore reduces priced-in risk — but the tender process will still produce its own set of questions.
Is a buildability review the same as a constructability review? In UK practice the terms are used interchangeably. "Constructability" is more common in US literature; "buildability" is the term you'll see in RIBA and CIOB documents.
The takeaway
A buildability review isn't about a contractor second-guessing an architect's design. It's about surfacing the dozens of small decisions where your drawing and the physical build will disagree, and deciding now, on screen, rather than later, on site.