Article 5: Factors Affecting Estimates and Cost Accuracy
Introduction
Every construction estimate is a prediction. It answers the question: what will this building cost, built in this location, to this specification, by this procurement method, at this point in time? Change any one of those variables and the answer changes with it. The same 24-apartment residential block that costs £3,950,000 in Salford could cost £5,020,000 in London or £3,800,000 if built to a budget specification. A six-month programme delay can add £600,000. A switch from traditional procurement to design and build might add 3–5%.
Understanding what drives these differences — and how to quantify them — is one of the most important skills a quantity surveyor can develop. It is the difference between an estimate that survives contact with reality and one that becomes a source of embarrassment when tenders arrive.
This article — the fifth in the estimating series — examines the key factors that affect construction cost estimates: location, market conditions, procurement route, specification, programme, ground conditions, and the inherent uncertainty in all estimating work. It explains why two competent QS professionals pricing the same project can produce figures 10–15% apart, and what the profession can do about it. Throughout, Project Parkside is used to show how each factor changes the number.
Location: Where You Build Changes What You Pay
Location is the single largest external factor affecting construction cost. The same building designed to the same specification will cost significantly more in London than in the North East, primarily because of labour rates, site logistics, and the local regulatory environment.
The BCIS provides standardised location adjustment factors that allow the QS to translate cost data from one region to another. These indices are updated quarterly and reflect actual tender price data from across the UK.
| Region | BCIS Index (SE = 100) | Variation from SE Base |
|---|---|---|
| London | 115–120 | +15% to +20% |
| South East | 100 | Base (0%) |
| South West | 94–97 | –3% to –6% |
| Eastern | 96–98 | –2% to –4% |
| East Midlands | 92–95 | –5% to –8% |
| West Midlands | 91–94 | –6% to –9% |
| Yorkshire and Humber | 89–93 | –7% to –11% |
| North West | 88–92 | –8% to –12% |
| North East | 86–90 | –10% to –14% |
| Wales | 89–93 | –7% to –11% |
| Scotland | 90–95 | –5% to –10% |
| Northern Ireland | 88–92 | –8% to –12% |
The variation is driven primarily by labour cost. An experienced bricklayer in London commands £180–£220 per day (self-employed rate) compared to £140–£170 in the North West and £120–£150 in parts of the North East. Specialist trades — particularly MEP and structural steelwork — carry premiums of 10–20% in high-demand areas. Material costs are more uniform nationally (delivered prices vary by only 3–8% between regions), but urban site logistics add a premium: restricted delivery hours, congestion charges, limited storage, and the need for just-in-time scheduling all increase the cost of getting materials to the work face.
Local planning requirements also matter. Section 106 contributions in growth areas can add £5,000–£20,000 per residential unit. Community Infrastructure Levy rates vary dramatically by local authority — Salford charges approximately £60–£90/m² for residential, while some London boroughs charge over £200/m². These costs are typically borne by the developer rather than the contractor, but they form part of the total project budget that the QS must track.
Project Parkside: Location Comparison
| Location | BCIS Index | Estimated Tender | Difference from Salford |
|---|---|---|---|
| Salford, North West (baseline) | 92 | £3,950,000 | — |
| Edinburgh, Scotland | 93 | £3,993,000 | +£43,000 (+1.1%) |
| Birmingham, West Midlands | 93 | £3,993,000 | +£43,000 (+1.1%) |
| Bristol, South West | 96 | £4,122,000 | +£172,000 (+4.4%) |
| London | 117 | £5,024,000 | +£1,074,000 (+27.2%) |
The London figure deserves emphasis: the same building, same specification, same number of apartments — but £1,074,000 more expensive. That is almost entirely a labour and logistics premium. The materials going into the building are broadly the same price; it is the cost of the people installing them that changes.
Market Conditions and Tender Price Inflation
Construction costs do not stand still. The BCIS Tender Price Index (TPI) tracks the movement of tender prices over time, and it tells a dramatic story. From 2020 to 2023, the TPI rose approximately 35% — driven by COVID-19 supply chain disruption, the energy price spike of 2022–23, and the Brexit-related reduction in available EU labour. By 2025–26, inflation has settled to 3–3.5% per annum, but this is still significantly above the pre-2020 norm of 0.5–2%.
Material Price Volatility
Materials are the most volatile cost component. The QS must understand which materials are subject to commodity-driven price swings and allow for them in the estimate.
| Material | 2021 Peak | Current (Q1 2026) | Volatility |
|---|---|---|---|
| Structural steel (reinforcement) | £1,100/tonne | £650–£750/tonne | ±5–10% quarterly swings; stabilised but energy-sensitive |
| Softwood timber | £1,200/m³ | £550–£650/m³ | High — global shipping, UK sawmill capacity |
| Copper (cable, pipe) | £9,500/tonne | £8,800–£9,200/tonne | Very high — commodity-driven; energy transition demand |
| Ready-mix concrete (C30) | £130–£150/m³ | £140–£165/m³ | ±10–15% seasonal; energy-intensive production |
| Facing bricks | £500–£700/1,000 | £600–£800/1,000 | Low — gradual 2–3% p.a. increase; lead times 8–12 weeks for specials |
The practical implication for the QS is that an estimate based on Q1 2026 material prices may be 3–5% out of date by Q3 2026 if commodity prices move. For projects with a long pre-construction phase, the QS should apply a tender price inflation allowance — typically based on the BCIS TPI forecast — to bring the estimate forward to the anticipated tender date.
The Economic Cycle
Construction costs are cyclical. In an expansion phase (rising workload, contractor confidence, tight labour market), tender prices rise as contractors can be selective and price in higher margins. In a contraction phase (falling workload, recession risk), tender prices fall — but with a 6–12 month lag, because labour rates are sticky downward and material suppliers maintain prices while demand softens. The 2008–12 financial crisis saw the TPI fall 16% from peak to trough, with recovery taking eight years. The QS who understands the cycle can advise the client on the optimal timing for procurement — going to tender in a soft market can save 5–15% compared to peak conditions.
Brexit has left a permanent mark on the UK construction labour market. Pre-2016, approximately 35% of skilled construction workers were EU nationals. By 2026, that figure has settled at 10–15%. The result is a structural wage uplift of 5–10% for key trades — a cost that is now baked into all estimates and is unlikely to reverse.
Procurement Route
The method by which a project is procured affects both the price and the certainty of the estimate. Different procurement routes allocate risk differently between the client and the contractor, and this allocation is priced.
| Procurement Route | Estimated Cost | vs. Traditional | Cost Certainty | Best For |
|---|---|---|---|---|
| Traditional (JCT Standard) | £3,950,000 | Baseline | High | Standard projects; competitive market |
| Design and Build | £4,080,000 | +3.3% | Medium | Fast-track; performance specifications |
| Two-Stage Tender | £4,030,000 | +2.0% | Medium–High | Complex projects; early contractor input |
| Construction Management | £4,090,000 | +3.5% | Very Low | Sophisticated clients; maximum flexibility |
| Management Contracting | £4,180,000 | +5.8% | Low | Complex projects; early start needed |
| Framework Agreement (Year 2+) | £3,880,000 | –1.8% | High | Repeat clients; portfolio programmes |
| Negotiated Contract | £4,200,000 | +6.3% | Low | Sole supplier; urgent; continuation work |
The key trade-off is between price and certainty. Traditional procurement through competitive lump sum tendering typically produces the lowest headline price because competitive tension is strongest — three to five contractors bidding for the same job drives prices down. But the price is only certain at the point of tender; variations, claims, and unforeseen conditions can erode that certainty during construction.
Design and build adds 3–5% because the contractor is carrying design risk and prices conservatively to protect their margin. But it offers faster procurement (the contractor can start construction while design is still being finalised) and single-point accountability. Framework agreements can produce the lowest costs of all — but only after the first year, once the relationship efficiencies offset the initial setup premium.
For the QS, the choice of procurement route is not just a cost decision — it determines how the estimate will be used. In traditional procurement, the QS’s cost plan is the benchmark against which tenders are assessed. In design and build, the QS’s cost plan becomes the employer’s requirements budget. In management contracting, the QS’s role shifts from estimator to cost controller, monitoring actual expenditure against forecast in real time.
Specification and Design Complexity
Specification drives cost more directly than any other factor within the design team’s control. The QS’s role is to help the client understand the cost implications of specification choices — and to quantify the trade-offs so that decisions are informed, not arbitrary.
Specification Impact
| Specification Level | Cost per m² GIFA | Project Parkside Total |
|---|---|---|
| Budget (standard brick, PVCu windows, vinyl flooring, gas boiler, basic kitchens) | £1,900/m² | £3,800,000 |
| Standard (quality brick, aluminium windows, engineered wood, condensing boiler, fitted kitchens) — baseline | £2,050/m² | £4,100,000 |
| Premium (premium brick, triple-glazed aluminium, solid oak, heat pump, bespoke kitchens) | £2,350/m² | £4,700,000 |
The gap between budget and premium specification is £900,000 — a 24% uplift — on the same building footprint. The biggest single contributors are the heating system (heat pump vs gas boiler: +£30,000–£50,000), windows (triple-glazed vs PVCu: +£40,000–£80,000), and kitchens (bespoke vs basic: +£15,000–£40,000 per apartment).
Design Complexity
Geometry matters. A simple rectangular floor plan with a regular structural grid is the cheapest form to build because it maximises repetition, simplifies formwork, and allows standard component sizes. Irregular footprints, curved facades, and complex roof forms add cost through bespoke formwork, specialist subcontractors, longer design coordination, and reduced productivity on site. A moderately complex design (L-shaped footprint, mixed uses) typically adds 5–10% to construction cost; a highly complex design (curved facades, multiple materials, complex geometry) can add 15–25%.
Repetition is the QS’s best friend. On Project Parkside, 24 apartments arranged in four similar floor plates means the labour force builds the same layout six times per floor. By the third repetition, labour productivity has improved by 15–20% compared to the first — a cumulative saving of £240,000–£300,000 over the project. This is why standardisation is such a powerful cost management tool, and why the QS should actively encourage design teams to maximise repetition wherever the brief allows.
The Cost of Late Design Changes
| RIBA Stage | Typical Cost Impact of Change | Example | Estimated Cost |
|---|---|---|---|
| Stage 2 (Concept) | ±5% of element cost | Switch from timber frame to masonry | +£80,000–£120,000 |
| Stage 3 (Developed) | ±8–12% of element cost | Upgrade windows to triple glazing | +£40,000–£80,000 |
| Stage 4 (Technical) | ±10–15% of element cost | Change structural system | +£150,000–£300,000 + 8–12 week delay |
| Stage 5 (Construction) | ±15–25% of element cost | Late kitchen specification change | +£40,000 direct + £50,000–£100,000 delay |
The pattern is clear: design changes become exponentially more costly as the project progresses. A specification change that costs £50,000 to implement at concept design might cost £150,000–£250,000 during construction, once disruption, programme impact, and extended preliminaries are factored in. This is why the QS should push for early design freeze and resist late changes unless they are genuinely necessary.
Programme and Time
Time is money in construction — literally. Preliminary costs are predominantly time-related: site management salaries, plant hire, scaffolding, temporary services, welfare, and security all accrue weekly or monthly. On Project Parkside, time-dependent preliminaries run at approximately £67,000 per month. Every month the programme extends adds £67,000 to the cost — regardless of whether any productive work is being done.
This makes programme duration one of the most significant cost drivers on any project. A six-month delay to Project Parkside (from 18 months to 24 months) adds approximately £400,000 in extended preliminaries alone — before accounting for material and labour inflation over the additional period. The total cost of a six-month delay, including inflation, is approximately £615,000 — a 15.6% uplift on the tender price.
Acceleration works in reverse but is not symmetrical. Compressing the programme from 18 to 12 months saves £400,000 in preliminaries but incurs acceleration costs — double-shift working (15% labour uplift), material expediting, and additional temporary works — that can total £400,000–£560,000. The net effect is usually a cost increase, not a saving. The rule of thumb is that reducing programme by 30% typically costs 8–15% in total project uplift.
Seasonal working adds a further layer. Winter working (November to February) carries a 5–10% cost premium due to frost protection for concrete, reduced daylight hours, lower labour productivity, and energy costs for drying and heating. The QS should factor seasonal timing into the estimate, particularly for groundworks and superstructure phases that are most exposed to weather.
Ground Conditions and Site Factors
Ground conditions are the greatest source of unforeseeable cost on most construction projects. What lies beneath the surface — the bearing capacity of the soil, the depth of the water table, the presence of contamination, the existence of buried services — can transform the cost of a project’s foundations from a manageable 7–9% of structural cost to a project-altering 25–30%.
Foundation type is the primary cost driver. Project Parkside’s baseline assumes pad foundations on London Clay at 0.8–1.2m depth — a cost of approximately £300,000. If the ground investigation reveals inadequate bearing capacity and piling is required, the foundation cost rises to £1,400,000–£1,500,000 — an uplift of over £1,000,000. This is why the QS must understand the ground investigation report and price foundation risk explicitly, not bury it in a blanket contingency.
Contamination is the other major ground risk. Light contamination (hydrocarbon traces, low-level heavy metals) typically costs £50,000–£200,000 to remediate. Moderate contamination (solvents, PCBs) can cost £200,000–£800,000. Severe contamination (asbestos, severe hydrocarbon) can exceed £1,000,000 and may render the project unviable within the original budget. Phase 1 and Phase 2 environmental assessments are essential pre-purchase diligence — and the QS should insist that their results are available before finalising the cost plan.
Other site factors include water table depth (dewatering during construction: £50,000–£120,000), flood risk (mitigation measures: +3–7% of project cost), site topography (sloping sites: +4–8%), and the presence of existing structures requiring demolition (£10,000–£50,000 for minor buildings; £280,000–£1,050,000 for substantial structures with asbestos).
Estimating Accuracy and Uncertainty
Every estimate has an accuracy range, and that range narrows as design develops. Understanding this — and communicating it clearly to clients — is one of the QS’s most important responsibilities.
| RIBA Stage | Estimate Type | Accuracy Range | Project Parkside Cost Range |
|---|---|---|---|
| Stage 1 (Preparation) | Order of magnitude | ±25–35% | £2,960,000–£5,190,000 |
| Stage 2 (Concept) | Elemental estimate (CP1) | ±15–20% | £3,160,000–£4,620,000 |
| Stage 3 (Developed) | Detailed elemental (CP2) | ±8–12% | £3,470,000–£4,410,000 |
| Stage 4 (Technical) | Pre-tender estimate (CP3) | ±5–8% | £3,620,000–£4,225,000 |
| Stage 5 (Tender) | Tender price | ±3–5% | £3,760,000–£4,085,000 |
The range at Stage 1 is enormous — £2,960,000 to £5,190,000 — because almost nothing is known about the design. By Stage 4, the range has narrowed to £3,620,000–£4,225,000, but even here there is a £600,000 spread. The QS who presents a single-point estimate without an accuracy range is doing the client a disservice: they are implying a level of precision that does not exist.
Optimism Bias
Research consistently shows that construction cost estimates are systematically optimistic. The HM Treasury Green Book guidance on project appraisal recommends applying an optimism bias adjustment of 10–20% to construction cost estimates, reflecting the finding that projects routinely exceed their budgets by this margin. The Mott MacDonald study (2002) and Oxford Brookes research (2014) both found that UK residential projects average 15–20% cost growth from estimate to outturn.
This is not a reflection of poor estimating. It is a systemic characteristic of construction projects: designs evolve, site conditions surprise, market conditions shift, and clients change their minds. The QS who understands optimism bias can advise clients to set budgets at P80 (the 80th percentile of likely outcomes) rather than the deterministic estimate, providing a realistic buffer for the variance that experience shows will occur.
Reference Class Forecasting
An alternative to bottom-up estimating is reference class forecasting — predicting the cost of a project based on the actual outturn costs of similar completed projects, rather than building an estimate from first principles. For Project Parkside, a reference class of 20–30 comparable apartment schemes completed in 2023–2025 produces a statistical forecast of approximately £4,100,000 (±£360,000) — 3.8% above the bottom-up estimate of £3,950,000. The difference captures the systematic optimism that bottom-up estimates tend to contain. Best practice is to use both methods: where they diverge by more than 5%, the assumptions driving the difference should be investigated.
Risk and Contingency Management
Contingency is not a slush fund — it is a quantified allowance for identified risks. The QS’s job is to reduce contingency over the life of the project by progressively resolving the uncertainties it covers, not to carry it unchanged from concept to tender.
| RIBA Stage | Total Contingency | Design Risk Component | Construction Risk Component |
|---|---|---|---|
| Stage 1 (OOM) | 25–35% | 20–30% | 5–10% |
| Stage 2 (CP1) | 15–20% | 12–18% | 3–8% |
| Stage 3 (CP2) | 8–12% | 6–10% | 2–4% |
| Stage 4 (CP3) | 5–8% | 3–5% | 2–4% |
| Stage 5 (Tender) | 3–5% | 0–2% | 3–5% |
The pattern is clear: design risk dominates early (when design is uncertain) and diminishes as the design is developed and frozen. Construction risk remains relatively constant — there are always risks associated with delivery, regardless of how well the design is documented. The QS’s role is to ensure that contingency reflects actual risk exposure, not just a comfortable round number. A contingency of “10%” at Stage 3 should be decomposable into specific items: £X for ground condition uncertainty, £Y for specification development, £Z for market risk.
For high-value or high-risk projects, Monte Carlo simulation can be used to model contingency probabilistically. Rather than a single deterministic estimate, the QS assigns probability distributions to key uncertain variables (labour productivity, material costs, subcontractor rates) and runs thousands of iterations to produce a probability curve. For Project Parkside, a Monte Carlo analysis might show: P50 (median outcome) £4,050,000, P80 (budget confidence) £4,200,000, P95 (high confidence) £4,350,000. This gives the client a much richer picture of risk than a single number with a contingency percentage.
Project Parkside: How Factors Change the Number
The table below brings together the key factors and shows how each one moves the Project Parkside estimate from the baseline tender of £3,950,000.
| Factor | Scenario | Revised Estimate | Impact |
|---|---|---|---|
| Location | London instead of Salford | £5,024,000 | +£1,074,000 (+27%) |
| Procurement | Design and build instead of traditional | £4,080,000 | +£130,000 (+3.3%) |
| Specification | Premium instead of standard | £4,700,000 | +£750,000 (+19%) |
| Programme | 6-month delay | £4,565,000 | +£615,000 (+15.6%) |
| Ground conditions | Piling required instead of pad foundations | £5,000,000 | +£1,050,000 (+27%) |
| Market timing | 12-month procurement delay (3% TPI) | £4,069,000 | +£119,000 (+3%) |
| Framework procurement | Year 2+ framework agreement | £3,880,000 | –£70,000 (–1.8%) |
| Budget specification | Budget instead of standard | £3,800,000 | –£150,000 (–3.8%) |
The message is stark. Location and ground conditions can each add 27% to the baseline. Programme delay adds 16%. Specification adds 19%. These are not marginal effects — they are project-shaping. The QS who fails to identify and quantify these factors early is not managing cost; they are hoping for the best.
Why Two Estimators Can Produce Different Figures
It is entirely normal for two competent QS professionals to produce estimates for the same project that differ by 8–15%. This is not a failure of the profession — it reflects the inherent judgement involved in estimating and the legitimate range of assumptions that can be made about any project.
The main sources of legitimate disagreement are comparable project selection (which historical projects the QS uses as benchmarks: ±2%), specification assumptions (how the QS interprets the brief: ±3–8%), productivity factors (how efficient the QS assumes the contractor will be: ±2–5%), subcontractor rates (current market vs historical averages: ±2–5%), preliminaries assessment (the QS’s view of programme duration and site complexity: ±2–4%), contingency philosophy (risk appetite and stage of design: ±1–3%), and market outlook (the QS’s view of tender price inflation and competition: ±3–8%).
Cumulatively, these differences can be substantial. Two estimates for Project Parkside — one conservative at £4,534,000 and one optimistic at £3,878,000 — represent an 8.5% variance from the midpoint. Both are defensible. The conservative estimate assumes premium specification, higher preliminaries, and 8% contingency. The optimistic estimate assumes standard specification, leaner preliminaries, and 5% contingency. Neither is wrong; they reflect different professional judgements about the same project.
The best response when estimates diverge is not to pick the lower number (which clients instinctively prefer) but to identify the drivers of variance, explain the reasoning behind each assumption, present the range as the realistic budget envelope, and recommend validation through competitive tender or BCIS benchmarking. The QS who presents a range with explanation serves the client far better than the one who presents false precision.
Practical Tips
Always present accuracy ranges. A single-point estimate at any stage before tender is misleading. The client needs to understand not just the number but the confidence level behind it. At Stage 2, say “£3,840,000 ±15–20%”; at Stage 4, say “£4,103,000 ±5–8%”. This is not hedging — it is professional honesty.
Apply location adjustments rigorously. Never use benchmark data from one region without adjusting for the project’s actual location. A £/m² rate from a London project applied to a North West site will overstate cost by 20–30%. BCIS location indices exist for exactly this purpose.
Price ground risk explicitly. Do not bury foundation uncertainty in a blanket contingency. If the ground investigation is incomplete, state the assumption (“pad foundations assumed; if piling required, add £1,050,000”) and flag it as a risk item. This protects both the QS and the client.
Track material price movements. For projects with a long pre-construction phase, refresh key material prices (steel, concrete, copper, timber) quarterly. The BCIS TPI forecast and supplier quotations are the best sources. An estimate based on 6-month-old material prices is already out of date.
Quantify the cost of programme delay. Every client underestimates how expensive delay is. Calculate the monthly cost of time-dependent preliminaries and present it explicitly: “every month of programme extension costs £67,000 in additional site running costs.” This focuses minds on decision-making speed and design freeze.
Use reference class forecasting as a sense-check. After completing a bottom-up estimate, compare it to the outturn costs of similar completed projects. If the bottom-up figure is more than 5% below the reference class average, investigate the assumptions. The gap may reflect genuine project-specific savings — or it may reflect optimism bias that needs correcting.
What’s Next
The final article in this series — Technology in Estimating — looks at how digital tools are transforming the estimating process. It covers digital takeoff, BIM-based quantity extraction, estimating software platforms, and the use of cost data analytics in modern QS practice.
Series Cross-References
Article 1: Introduction to Estimating in Construction — what estimating is, why it matters, and how estimates evolve through the project lifecycle.
Article 2: Types of Estimates and When to Use Them — from order-of-magnitude through to detailed estimates, with the full Project Parkside cost evolution.
Article 3: Elemental Cost Planning (Client-Side) — the NRM 1 framework, BCIS benchmarking, and how the QS builds cost plans from CP1 through CP3.
Article 4: Contractor Estimating and Tender Pricing — first-principles rate build-ups, all-in labour rates, subcontractor procurement, and the tender process.
Article 6: Technology in Estimating — digital takeoff, BIM integration, estimating software, and the future of cost data. (Coming next)
External References
BCIS — Building Cost Information Service — the primary source of location indices, tender price indices, elemental analyses, and regional cost factors.
RICS NRM Standards — the New Rules of Measurement framework for cost estimation and cost planning.
RIBA Plan of Work 2020 — the design stage framework that defines the accuracy expectations at each stage.
HM Treasury Green Book — guidance on optimism bias adjustments for construction project appraisal.