The Three Line Items That Quietly Kill Capital Project ROI

When a capital project underperforms, the post-mortem usually focuses on the wrong numbers.

The change orders get audited. The scope creep gets blamed. Someone produces a chart showing budget variance by trade and the conversation moves to whether the GC was the right pick. Meanwhile, the line items that actually destroyed the return on the project are sitting in the closeout documentation, unexamined, because they never showed up in the headline variance.

Three of them, in particular, do most of the damage. They erode ROI quietly, month after month, while everyone’s attention is on the parts of the budget that move in dramatic amounts. By the time the project is finished, they’ve taken more off the return than any single change order ever did.

1. General conditions extensions

General conditions are the contractor’s cost of being on site. Supervision, trailer, dumpsters, temporary utilities, project management staff, small tools. On most projects they run somewhere between 8 and 14 percent of construction cost. They’re priced as a function of duration.

Which means every month of schedule extension carries a general conditions bill, regardless of whether any actual construction work happened in that month.

This is where ROI bleeds out of a project without anyone calling it a loss.

A six-month schedule extension on a $40M build can carry $1.5M to $3M in extended general conditions alone. That’s before any change order for the work that caused the delay. The owner pays the contractor to stand on site and supervise nothing, because the contract says they do.

The trap is that general conditions extensions rarely arrive as a single, debatable invoice. They show up as a monthly application for payment that’s technically aligned with the contract. The first month nobody objects, because the delay is fresh and the responsibility is unclear. The second month sets the pattern. By month four, the extended general conditions are just a line on the pay app, and the cumulative cost is buried in the totals.

There are two moves that protect the owner here. First, establish at the start of any delay event which party is on the hook for time-related costs, in writing, before the first extended pay app gets submitted. Second, require the contractor to substantiate extended general conditions as actually incurred. Not as a daily rate multiplied by delay days. As documented spending on supervision and site costs during the extended period. That distinction often cuts the bill significantly, because not every general conditions line item runs at full burn during a delay.

2. Soft cost burn during owner-caused delays

Hard costs get the attention because they’re contractually visible. The soft costs accumulate quietly. Design fees, owner’s representative fees, legal, project management, financing carry, insurance premiums tied to construction period. They rarely get aggregated until closeout.

On a complex project, soft costs run 15 to 30 percent of total project cost. During an extension, most of them keep running.

The architect of record is still billing for site visits and submittal reviews. The owner’s project manager is still on payroll. The MEP designer is still answering RFIs. The lender is still charging interest on the construction loan. Builder’s risk and other policies tied to the construction period are still in force, and may need to be extended at premium rates if the original term expires.

None of this is anyone’s fault, exactly. It’s the cost of an open project. The problem is when the cause of the extension is owner-driven. A late design release. An unresolved permit. A procurement decision that wasn’t made on time. Those soft costs aren’t recoverable from anyone. They come straight off the project return.

The number that surprises owners at closeout is often the soft cost overrun, not the hard cost overrun. The hard cost overrun was tracked. The soft cost overrun accumulated in nine different vendor relationships, each one a few thousand dollars over, and the total only became visible when the accountants closed the books.

The discipline that prevents this is unglamorous. Track soft cost burn against schedule on a monthly basis, the same way hard costs get tracked. Run a soft cost forecast every month that asks: if this project finishes three months late, what’s the additional soft cost exposure? Most owners don’t do this because soft costs feel like overhead, not project cost. By closeout they realize the distinction was expensive.

3. Contingency drawdowns that get rebranded

Every capital project carries contingency. Owner’s contingency, design contingency, construction contingency, sometimes all three. They exist to cover the unknowns that surface during execution.

The problem isn’t that contingency gets used. The problem is what gets called contingency.

In the first third of a project, contingency draws tend to be legitimate. Unforeseen conditions. Design clarifications. Scope refinements that fall within reasonable interpretation of the documents. By the middle of the project, the character of the draws shifts. Things start getting paid out of contingency that should arguably have been paid out of the original budget or recovered from another party. Late design changes the design team should have absorbed. Trade work that was clearly missed in the original buyout and could have been negotiated harder. Owner-initiated changes that nobody wanted to call a change order because it would be politically uncomfortable.

By the last third of the project, contingency is functioning as a slush account. It’s where uncomfortable costs go to disappear.

This matters for ROI because contingency was supposed to be the buffer that allowed the project to absorb genuine unknowns without compromising the return. When it gets drawn down for things that aren’t actually unknowns, two things happen. The first is that when a real unknown shows up at month 20, the contingency is gone. The second is that the project return looks like it landed within budget, when in fact it landed within budget plus contingency. From an investor’s perspective, that’s a very different number.

The fix is procedural. Every contingency draw should require a justification that names the category. Unforeseen condition. Design refinement. Owner-directed scope change. Recovery from another party. The justification should be reviewed by someone outside the project team. When that review exists, the late-stage rebranding stops. When it doesn’t, contingency becomes whatever the project team needs it to be at the moment.

The reason these three hide

None of these line items hide because anyone’s trying to hide them. They hide because the project’s reporting structure isn’t built to surface them.

Monthly project reports tend to focus on hard cost performance against budget. Schedule reports focus on critical path activity. Change order logs focus on documented changes. None of these reports naturally show extended general conditions burn rate, soft cost run-rate against schedule, or the qualitative shift in what contingency is being used for.

A reporting framework that surfaces these requires deliberate design. It isn’t a standard P6 export. It’s the kind of thing that needs to be built into the project controls plan at the beginning of the project, because adding it in the middle means trying to reconstruct nine months of data that wasn’t being captured.

When the framework exists, the conversation at the monthly owner meeting changes. Instead of reviewing where hard costs landed versus the budget, the owner can see how much money the schedule is costing them per month, where soft costs are tracking, and whether contingency is being drawn down at a rate consistent with the project’s stage. The hidden line items stop being hidden.

When the framework doesn’t exist, ROI quietly compresses, and the explanation only arrives at closeout. By then there’s nothing to do but write the post-mortem and try to apply the lesson to the next project.

Capital projects rarely fail dramatically. They fail in increments small enough that nobody is measuring closely enough to flag them. The owners who protect their returns are the ones who insist on measuring those increments from the beginning. Especially when the project is going well, because that’s when the early signs of all three of these are easiest to spot.