Construction Delay Statistics: What the Research Says and What It Means for Your Projects
By Kush Agarwal, President and CEO, YelloSKYE. Published .
Executive Summary
- The Question: How often do construction projects finish late, how reliable are the figures people quote, and what can a project team actually do about it?
- The Answer: The major studies all show that large projects usually finish late and over budget, though their exact numbers measure different things. The causes are well known. What separates teams that recover from those that do not is how quickly a slip is seen and acted on.
- The YelloSKYE Angle: Construct IQ gives project teams a dated visual record of the site, compared against the programme in DEX, so slippage shows up in the week it starts instead of at the month-end report.
Most large construction projects finish late. The best-known research, from McKinsey, KPMG and the Oxford-based project researcher Bent Flyvbjerg, puts the share of large projects finished on time somewhere between about a quarter and fewer than one in ten, depending on what is being counted. Delay is the expected outcome, and planning that assumes otherwise is planning for an exception.
This guide sets out the figures that are worth quoting, explains why they differ, and looks at the causes behind them. It then covers the part a project team controls: how fast a slip is noticed. That is where site visibility, such as the dated captures YelloSKYE delivers through Construct IQ, makes its difference. For the wider practice of tracking progress, see our construction progress monitoring guide.
The figures worth quoting
Plenty of numbers circulate about construction delays, many of them repeated without a traceable source. The ones below come from named organisations and are widely cited in the industry. Each measures something slightly different, so read the third column before comparing them.
| Finding | Source and year | What was measured |
|---|---|---|
| Large projects typically take about 20 per cent longer than scheduled and can run up to 80 per cent over budget | McKinsey Global Institute, Reinventing Construction, 2017 | Large projects across asset classes |
| 77 per cent of megaprojects are at least 40 per cent late, and 98 per cent overrun their budget by more than 30 per cent | McKinsey & Company, The construction productivity imperative, 2015 | Megaprojects only |
| Only 25 per cent of projects finished within 10 per cent of the original deadline, and 31 per cent within 10 per cent of budget | KPMG, Global Construction Survey, 2015 | Project owners reporting on their own projects over the previous three years |
| 8.5 per cent of projects were delivered on time and on budget; 0.5 per cent also delivered the benefits promised | Bent Flyvbjerg and Dan Gardner, How Big Things Get Done, 2023 | A database of more than 16,000 large projects of many types, not only construction |
| Global construction labour productivity rose roughly 1 per cent annually across 20 years; the world economy managed 2.8 per cent and manufacturing 3.6 per cent | McKinsey Global Institute, Reinventing Construction, 2017 | The global construction sector |
The last row is not a delay figure, but it explains part of the pattern. An industry whose output per worker has barely moved in twenty years has little slack to absorb surprises, so a lost week is rarely won back.
Why the numbers disagree
A figure of "25 per cent on time" and one of "77 per cent at least 40 per cent late" can both be right, because they answer different questions.
- Project size. Megaprojects fare worst. A study limited to them will show far higher overruns than one that includes ordinary buildings.
- The baseline. Some studies measure against the programme agreed at contract award; others use the latest approved revision. A project that has had its end date moved twice can be "on time" against the second baseline and a year late against the first.
- Who answers. Owner surveys report the owner's view of lateness. Contractors, who may have secured extensions of time, often see the same project differently.
- What counts as late. A 10 per cent tolerance, a 40 per cent threshold and "any slip at all" will produce very different headline percentages.
When you quote a figure in a board paper or a bid, quote the source, the year and what it measured. A number stripped of those three things is easy to dismiss.
Why a late finish costs more than the extra days
Delay is often discussed as a schedule problem, but every party pays for it in a different currency. That is why late projects so often turn into disputes: the costs are real, they land unevenly, and each side wants to pass them on.
| Who carries it | Typical costs of delay | How it usually surfaces |
|---|---|---|
| Main contractor | Prolonged site preliminaries, plant and scaffold hire, retained supervision, acceleration costs, liquidated damages if the delay is its responsibility | In the final account and any claim for an extension of time |
| Subcontractors | Idle or split crews, remobilisation, work moved into worse conditions or a less efficient sequence | Disruption claims, often hard to prove |
| Owner or developer | Extra finance costs, later revenue or rent, carrying costs on the land, internal management time, knock-on effects on sales or tenants | Overrun against the development appraisal |
| Lender or investor | Drawdowns out of line with progress, a longer exposure period, covenants under pressure | Monitoring reports and loan renegotiation |
The figures in this table are the ones that make a slip matter. They also explain why the evidence behind any delay claim becomes so valuable: whoever can show what happened on site, and when, is far better placed to settle who pays. Our guide to construction dispute documentation covers what that evidence needs to look like.
What causes construction delays
Researchers have studied delay causes on projects in dozens of countries, usually by surveying owners, consultants and contractors and ranking the factors they name. The rankings vary with the market and the type of project, but the same groups appear again and again.
| Cause group | Typical examples | Can a visual site record help? |
|---|---|---|
| Owner decisions and payments | Late approvals, slow responses to queries, delayed payment to the contractor | Yes. Shared, current evidence shortens the time to a decision and supports prompt payment |
| Design and information | Design changes, errors and omissions, late drawings, unresolved queries from site | Partly. It shows where built work departs from the design early, before more is built on top |
| Contractor planning and site management | Unrealistic programme, poor sequencing, weak supervision, trades getting in each other's way | Yes. It shows zones falling behind and work out of sequence while it can still be corrected |
| Labour, materials and plant | Shortages, late deliveries, low productivity, equipment breakdowns | Partly. It shows the effect on site quickly, though not the root cause in the supply chain |
| Finance and commercial issues | Contractor cash-flow problems, payment disputes, contested variations | Partly. Verified progress reduces the arguments that hold payments up |
| External conditions | Weather, permits, unexpected ground conditions, utility diversions | Rarely prevents them, but records their effect for any claim |
Two points stand out. First, most of the causes sit inside the project team rather than outside it. Weather and permits make headlines, but late decisions, changing information and poor coordination account for a large share of the lost time. Second, very few delays have a single cause. A late drawing leads to a crew moving to another area, which pushes a follow-on trade, which misses its window before a holiday shutdown. Each step is small; together they cost months.
The real problem: slips are found late
Almost every delay starts as something small and visible on site: a zone that has not moved for a week, a crew that is short-staffed, an element built in the wrong place. The question is how long it takes before anyone with authority sees it and acts.
On many projects the answer is weeks. Progress reaches management through a chain of summaries: a site engineer's notes, a weekly report, a monthly progress meeting. Each step adds delay and some optimism. Activities sit at "nearly finished" in the tracker for weeks, because the last part of a task is often the slowest and the hardest to see from a spreadsheet. By the time the programme shows a problem on the critical path, the float that might have absorbed it has gone.
Two kinds of lag
It helps to separate the delay you can see into two parts:
- Detection lag: the time between a slip starting on site and the project team recognising it.
- Decision lag: the time between the team recognising it and a decision being made, for example to resequence, add resources or issue an instruction.
Most projects only track the outcome, the days lost against the programme. Measuring these two lags tells you whether your reporting and your approvals are fast enough to recover lost time, which is the part you can improve.
How site visibility reduces delay
Visibility does not make a site build faster. It shortens both lags above, so that recoverable problems are caught while they are still recoverable. In practice that happens in five ways.
Seeing slippage against the programme
A capture is only useful for delay if it can be compared with the plan. In DEX, YelloSKYE overlays actual progress on the planned schedule, including 4D programmes, so deviations show up zone by zone rather than as one overall percentage. Leadership dashboards in DEX track stage-wise completion and lag across a whole portfolio. Our portfolio monitoring use case describes how capturing every site with the same drone, 360° and fixed-camera parameters let bottlenecks across a portfolio surface weeks earlier than manual reporting allowed.
Noticing when work stops
Some of the costliest delays are quiet ones: a zone where nothing happens for several days. Construct IQ sends a notification if activity in a high-priority area comes to a halt, and another when a critical milestone is reached. Fixed cameras keep a timestamped view of selected zones between scheduled drone flights, as in our fixed-camera construction monitoring work.
Making decisions without waiting for a site visit
Decision lag is often the time it takes for the right person to get to site. When the owner, project manager and consultants look at the same dated record remotely, many of those trips are no longer needed. Across verified deployments that pair fixed cameras with Construct IQ and DEX, YelloSKYE has recorded up to 80 per cent fewer site visits. In our aerial progress capture use case, stakeholder meetings went from two hours to 30 minutes because everyone arrived having seen the same site.
Catching errors before they are built over
Rework is one of the most common hidden causes of delay, because the time spent correcting an element is time not spent progressing the next one. Comparing laser scans or drone models against the design model after key stages finds misplaced elements while the fix is still quick. The cost side of this is covered in our guide to the cost of rework in construction.
Keeping evidence for extension-of-time claims
When delay happens despite everyone's efforts, the argument moves to who caused it. DEX keeps every capture in date order and lets reviewers scroll back through the site day by day, so the project team can show the state of the site at the moment a problem began. That record is also useful for an owner assessing a contractor's claim.
What visibility will not do
Visibility will not shorten a permit process, deliver a late shipment or stop a storm. It will not fix a programme that was never achievable in the first place. Its value is narrower and more practical: it removes the weeks that pass while a known problem waits to be noticed, believed and decided on.
Measuring delay on your own projects
Industry averages are useful context; your own figures are what you can improve. A short set of measures, reviewed monthly, will tell you more than any published survey.
| Measure | How to calculate it | What it shows |
|---|---|---|
| Milestone slip | Forecast date minus baseline date, in days, for each key milestone | Where the programme stands, against the original baseline as well as the current one |
| Float used | Float consumed on the critical path since the last review | How quickly the margin for recovery is disappearing |
| Activities started on time | Share of activities due to start in the period that actually started | Early warning, often weeks before a milestone moves |
| Detection lag | Days between a slip first being visible on site and it being reported | Whether reporting is fast enough to act on |
| Decision lag | Days between an issue being raised and a decision being made | Whether approvals, not the site, are the bottleneck |
| Delay by cause | Days lost, grouped by the cause categories above | Where to focus prevention on the next project |
Detection lag is the hardest to measure without a dated record of the site, because you need to know when the problem first appeared. With regular captures, you can go back to the first one that shows it.
A practical checklist
- Quote figures with their source. When using industry statistics in a business case, give the organisation, the year and what was measured.
- Keep the original baseline. Report slip against the first approved programme as well as the current one, so repeated rebaselining does not hide drift.
- Agree capture zones and cadence at mobilisation. Match drone flights, fixed cameras and 360° walks to the pace of each phase and to the payment cycle.
- Link every capture to the programme. A picture that cannot be tied to an activity cannot show delay.
- Set alerts on the zones that matter. Mark critical-path areas as high priority so that a stoppage is noticed in days, not at the next meeting.
- Track detection and decision lag. Review both monthly with the project manager and the owner's team.
- Record delay causes as they happen. A cause logged on the day, with the capture that shows it, is worth far more than one reconstructed at the final account.
- Review the archive at project close. Use the dated record to find where each major slip began, and feed that into the next project's programme and risk register, as described in our guide to construction risk management.
Frequently asked questions
What percentage of construction projects are delayed?
There is no single reliable percentage, because studies measure different kinds of project in different ways. The well-known surveys agree on the direction: KPMG's 2015 Global Construction Survey found only a quarter of owners' projects finished within 10 per cent of the original deadline, and McKinsey's research on megaprojects found more than three-quarters were at least 40 per cent late. On large projects, finishing late is the normal outcome rather than the exception.
What is the most common cause of construction delays?
Research from many countries points to the same group of causes rather than one: design changes and late information, slow decisions and payments by the owner, poor planning and site management by the contractor, shortages of labour, materials and equipment, and external factors such as weather, approvals and ground conditions. On most late projects, several of these act together, and small slips compound into large ones.
How much does a construction delay cost?
A delay costs more than the extra days of site overheads. The contractor carries prolonged preliminaries, idle plant and possibly liquidated damages, while the owner carries extra finance costs, delayed revenue and management time. Disputes over who caused the delay add legal and expert costs on top, which is why clear evidence of site progress matters so much.
Can technology reduce construction delays?
Technology cannot stop it raining or speed up a permit, but it can shorten the time between a slip starting and someone acting on it. Regular, dated site capture compared against the programme shows where work is falling behind while there is still float to recover it, and remote review cuts the wait for decisions that would otherwise need a site visit.
How do you measure delay on a construction project?
Start with the gap in days between planned and forecast completion for each key milestone, and track how much float has been used on the critical path. Add a measure of how long it takes your team to spot a slip after it begins, and how long decisions take once a problem is raised. Those last two show whether your reporting is fast enough to act on.