How Big Projects Lose Track Of Where They Actually Are

On a large capital project, “on schedule” is often a claim rather than a measurement. Someone reports a percentage complete at a weekly meeting, the number gets rolled up into a status report, and the report gets forwarded to people who were not on site to see whether it was true. By the time a discrepancy surfaces, weeks or months of work have already been built on top of an inaccurate picture.

This is not a failure of effort. Project teams on major builds are generally trying hard to stay current. The problem is structural: the systems most projects use to track progress were not built to catch drift between what is reported and what is actually in the ground, and the gap tends to widen exactly when a project is large, complex, or under schedule pressure.

Why the gap widens as projects get bigger

Construction productivity has been essentially flat for two decades. Global labor productivity in the industry rose only 10 percent between 2000 and 2022, an average of roughly 0.4 percent a year, compared with a 50 percent gain across the broader economy over the same period. From 2020 to 2022 alone, productivity in global construction actually declined by 8 percent, even as project complexity increased and brownfield work, which is harder to plan accurately, grew from 13 percent of the project pipeline in 2012 to 22 percent by 2022.

Technology adoption has not closed that gap, at least not yet. Investment in construction technology has grown substantially since 2020, but most of it has gone toward tools that increase control and reporting rather than tools that change how work actually gets measured and verified in the field. Digital document management and status tracking became common, but they largely digitized existing reporting habits instead of replacing subjective, self-reported progress figures with something that can be independently checked.

That distinction matters more as projects scale. On a small job, a superintendent can walk the site daily and know, with reasonable confidence, what has actually been built. On a hundred-million-dollar vertical build with multiple trades working in parallel across dozens of floors, that kind of informal verification breaks down. Status reports increasingly describe what teams believe is happening rather than what an independent record shows.

When the tracking system itself cannot track

The consequences of this gap are visible in some of the most heavily scrutinized capital programs in the country. A recent federal assessment of major nuclear infrastructure construction projects found that cumulative cost overruns across a portfolio of 28 major projects, each budgeted at more than $100 million, more than doubled in roughly two years, growing from $2.1 billion to $4.8 billion. Cumulative schedule delays across the same portfolio grew from nine years to thirty years over that period. Two projects alone accounted for about 80 percent of the total cost overrun.

The review attributed much of this to inadequate project management by contractors overseeing the work, along with vendor performance issues and rising input costs. Notably, these were not undocumented or informal projects. They had approved cost and schedule baselines, dedicated project management teams, and regular reporting requirements. The tracking infrastructure existed. What it lacked was a reliable way to verify, independent of self-reported status, what had actually been completed against that baseline.

Owners evaluating a construction progress monitoring approach for a large or multi-phase project are, in effect, trying to solve exactly this problem: replacing progress narratives with a record that can be checked against reality at any point, not just at scheduled milestones.

The paper trail that turns out not to be one

A recent municipal audit of a major public works agency offers a useful case study in how tracking gaps compound. The audit found that 80 percent of the agency‘s managed projects with completions after a certain fiscal year were behind their original schedule, with an average delay of 3.5 years. More striking than the delay itself was what the audit found about how that delay was tracked internally: the agency’s project management system did not retain the original cost estimate once a project proceeded, replacing it with updated figures as the project moved forward. This effectively erased the ability to measure total cost growth from the original baseline to completion.

The audit also documented a specific case in which a design consultant delivered a building information model at a level of development well below what the contract required, and construction proceeded before the gap was corrected. The incomplete model lacked sufficient detail in mechanical, electrical, and plumbing systems, and unresolved conflicts, including a boiler room without adequate space for its own equipment, surfaced only once construction was underway. The resulting delay and change orders traced directly back to a documentation standard that was accepted on paper but not actually met in practice.

In one sampled case, more than 60 percent of design-phase completions on a major public agency’s projects ran over 360 days late, and internal management reports understated construction delays by an average of 53 percentage points compared with the agency’s own underlying project data.

In both cases, a defined process existed on paper. The vulnerability sat one layer deeper: the documentation and reporting that fed that process could drift away from ground truth for months without anyone noticing, until the gap became too expensive to ignore.

What actually closes the gap

None of this suggests that more paperwork is the answer. If anything, both examples point toward the opposite conclusion: additional reporting requirements layered onto an already strained system tend to produce more reports, not more accuracy. What seems to matter more is the ability to verify status against an independent, dated record rather than relying entirely on a team’s own account of its progress.

A few practical implications follow from this:

Large projects will likely always carry some degree of uncertainty between planned and actual progress. Audits and reviews of troubled projects repeat a more specific pattern, though: the systems meant to catch that uncertainty were structured in ways that let a false sense of progress persist for months or years before anyone was forced to reconcile it with reality. Closing that gap depends less on adding oversight and more on making sure the oversight that already exists cannot be quietly overwritten.