Why cost control matters

A contractor's estimate is a prediction; its cost control system is how the prediction is turned into a result. Most general contractors and trade contractors work on thin margins. A building contractor that bids a 4 or 5 % fee on a lump sum contract can see the whole of that margin disappear with a small slip in labour productivity, a late subcontract buyout or a handful of unpriced changes. Cost control is the set of practices that catches those slips while there is still time to act: a budget the field understands, a way of recording what has been spent and committed, a way of measuring how much work has been done, and a regular forecast of where the job will finish.

This course follows a project from the moment the estimate becomes a budget to the monthly cost report that tells management and the owner where the job stands. It builds on CON-200 (estimating) and CON-210 (planning and scheduling), and uses the earned value ideas that CON-260 introduced as dashboard measures. Contract law for payment, changes and holdback is covered in CON-220; contractor accounting and revenue recognition belong to CON-320; procurement and buyout to CON-330; and risk, contingency and insurance to CON-350. Where those topics touch cost control, this course explains what the cost engineer needs and points to the course that goes deeper.

The evidence on cost overruns

Studies of large public projects have found for decades that most finish over budget. The research of Bent Flyvbjerg and colleagues on transport and infrastructure megaprojects is the best known: across thousands of projects, a large majority exceeded their budgets, and the average overrun for rail and bridge projects ran well above a quarter of the original estimate. Building projects tend to show smaller average overruns than megaprojects, but the spread is wide and a small share of projects produce very large losses. The causes are a mix of optimistic estimates, scope growth, design changes, poor site conditions, weak control and, on some projects, strategic misrepresentation when a low number is used to win approval.

Two lessons follow for the contractor. First, an overrun usually builds up gradually rather than appearing in a single event; the early months of the job give warning signs if anyone is measuring. Second, the cost of correcting a problem rises steeply with time. Research on project performance suggests that once a project is about 20 % complete, its cumulative cost performance rarely improves much, which is why early, honest measurement matters so much (Lesson 8 returns to this finding).

What cost control is, and what it is not

Cost control is not accounting, although it uses accounting data. Accounting records what has happened, by legal entity and period, to produce financial statements and pay bills. Cost control looks forward: it compares what has happened with what was planned and asks what will happen by the end of the job. The two systems share the job cost ledger, but the cost engineer adds three things the accountant does not need: a budget at the level of detail the field manages, a measure of physical progress, and a forecast.

Cost control is also not the same as cost cutting. Reducing spending on a job that is already well planned can make things worse; cutting a supervisor to save general conditions can cost far more in rework and delay. The aim is to know, early and accurately, where the job is heading, so that the project team can make good decisions.

The cycle in the figure is simple, but each step needs a discipline behind it. The budget must be broken into the same cost codes the field charges time and materials to. Actual cost must be recorded accurately and promptly, including costs that have been incurred but not yet invoiced. Progress must be measured objectively, not guessed. Variances must be explained by someone who knows the work. The forecast must reflect what the team really expects, not what it hoped at bid time. And action must follow, or the cycle becomes a reporting ritual.

Who does what

On a small project, the project manager may do all of this alone. On a large project, a project controls group may include cost engineers, schedulers and change managers. What matters is that every responsibility has an owner and that the field and office work from the same cost codes and the same budget.

How the contract type shapes cost control

The contract between owner and contractor determines who bears the risk of cost overruns, and so how the contractor's cost control is used and who sees the results.

Standard form contracts used in Canada, such as the CCDC stipulated price, cost plus and construction management forms, set out how payment applications, changes and cost records work; CON-220 explains them. Cost-reimbursable contracts usually give the owner audit rights over the contractor's records, so a cost system that cannot trace each dollar to a code, a date and a document becomes a contractual problem, not just an internal one.

The running example: Riverbend Commons

Throughout this course we follow Riverbend Commons, an invented four-storey mixed-use building with ground-floor retail and three floors of rental apartments, about 6,200 m² gross floor area. Northline Builders, the general contractor, won it under a stipulated price contract of $16,850,000 before taxes. Northline self-performs concrete formwork and placing, rough carpentry and general labour, and subcontracts the rest. Its bid carried a fee of 4.5 % and a contingency of 1.8 % of cost. The schedule runs 16 months.

The numbers in the example are illustrative and are chosen to make the calculations clear, not to represent current market prices.

How small slips erase a margin

Margin fade (or profit fade) is the gradual erosion of the expected profit on a job between the bid and the final accounting. It is a familiar pattern in contractors' financial results: each month the projected margin drops a little as cost problems surface. The simple calculation below shows why even modest overruns matter so much on a lump sum job.

contract = 16_850_000
fee_rate = 0.045
cost_budget = contract / (1 + fee_rate)          # cost the fee was applied to
planned_fee = contract - cost_budget
print(f"Cost budget {cost_budget:,.0f}, planned fee {planned_fee:,.0f}")
for overrun in (0.01, 0.02, 0.03, 0.045):
    final_cost = cost_budget * (1 + overrun)
    fee_left = contract - final_cost
    print(f"Overrun {overrun:.1%}: fee left {fee_left:>10,.0f}  "
          f"({fee_left / planned_fee:.0%} of planned)")

Cost budget 16,124,402, planned fee 725,598
Overrun 1.0%: fee left    564,354  (78% of planned)
Overrun 2.0%: fee left    403,110  (56% of planned)
Overrun 3.0%: fee left    241,866  (33% of planned)
Overrun 4.5%: fee left          0  (0% of planned)

A 2 % cost overrun takes away almost half of a 4.5 % fee, and an overrun equal to the fee rate leaves the job at break-even before any home office overhead is recovered. Because most of the cost on a building is subcontracted at fixed prices after buyout, the remaining risk sits in a smaller share of the work, chiefly self-performed labour, general conditions, unbought packages and changes. That is where the cost control effort should focus.

Reporting periods and the rhythm of control

Most contractors run a monthly cost cycle that follows the accounting close, with a formal cost report and forecast at each month-end. Labour is usually watched more often, weekly or even daily, because labour productivity can change quickly and is the main controllable cost on self-performed work. Large projects may have a weekly cost meeting where the superintendent, project manager and cost engineer review quantities installed and hours spent. The rhythm matters because a monthly report that arrives three weeks after month-end is reporting on a job that has moved on.

What good looks like

Good cost control systems share a few features regardless of size. The budget is set up before work starts and changes only through documented transfers or approved changes. Every cost is coded at the source, by the person who knows what it was for. Progress is measured by objective methods agreed in advance. The forecast is prepared by the project team, reviewed by someone independent, and compared with the previous month so that trends are visible. And the results feed back to estimating so the next bid is better.

The remaining lessons build those features one at a time. Lesson 2 turns the estimate into a control budget; Lesson 3 structures it with cost codes and loads it onto the schedule; Lessons 4 and 5 measure progress and record field costs; Lesson 6 assembles the cost report; Lessons 7 to 9 add earned value and forecasting; Lesson 10 deals with changes; Lesson 11 with cash; and Lesson 12 with action, reporting and a capstone monthly report.