According to FMI and Autodesk (2021), bad data cost the global construction industry an estimated €1.58 trillion in 2020. This is calculated by applying the IBM-estimated ratio of bad data costs to GDP (16.5%) to the construction sector's share of global GDP (13.2%). For a contractor with €1 billion in annual revenue, up to €165 million of revenue is estimated to be impacted by bad data, with €7.1 million in rework directly avoidable through better data quality.
Research & Evidence
The cost of
bad data
in construction.
Bad data is not a minor inconvenience. It is a structural cost that every construction company carries — embedded in rework, wrong decisions, and operational failures. The numbers are precise. The problem is solvable.
40%
of the average construction company's data is bad — inaccurate, incomplete, inconsistent, or untimely
41%
of the time, bad project data directly leads to poor decisions across European construction firms
€1.58T
estimated total cost of bad data to the global construction industry in 2020 alone
The evidenceConstruction generates more data
Construction generates more data
than ever. Most of it is unusable.
The research is unambiguous. Bad data in construction is not an edge case — it is the default. And its consequences extend far beyond individual projects.
A 2021 study by FMI and Autodesk surveyed 1,115 construction professionals across Europe. The findings are stark: 82% of firms are collecting more data than three years ago — yet 39% say less than half of that data is actually usable.
The root causes are consistent across every country surveyed. Data is inaccurate. Data is incomplete. Data is inconsistent between systems and parties. And critically, it is often untimely — arriving too late to influence the decisions it was meant to support.
The result: 41% of project managers make their decisions using data they cannot fully trust. In an industry where a wrong procurement decision or a misclassified element can cascade into days of rework, this is not a data quality problem. It is an operational risk.
The McKinsey Global Institute adds a longer-term dimension. Construction labor productivity has grown at just 1% per year over two decades — compared to 2.8% for the total economy. A significant part of that gap is attributable to information failures: wrong decisions made on bad data, repeated project by project.
"The level of decision making in the field has always been high, but these decisions have been made mostly on experience and not analysis."
BIM coordinator, German main contractor — FMI / Autodesk 2021
"We had a project that went terribly bad. Two years later, we had the exact same project. We made the same mistakes all over — everything was avoidable."
Innovation & quality manager, UK contractor — FMI / Autodesk 2021
By the numbersNine data points that
Nine data points that
define the problem.
14%
of all rework globally caused by bad data
Rework represents 5% of total global construction spend. Bad data decisions drive 14% of that — a disproportionate and largely preventable share.
€165M
impacted per €1B contractor annually
For a contractor performing €1 billion in work per year, upwards of €165 million of revenue is estimated to be affected by bad data across procurement, delivery, and operations.
95%
of construction data is never used
The vast majority of data captured on construction projects delivers no operational value — it is collected but never translated into decisions or downstream systems.
13%
of engineer time lost searching for data
Engineers and project managers spend over one in eight working hours simply locating information that should be immediately accessible — time that produces nothing.
30%
of construction costs are failure costs
Up to 30% of total construction project costs are attributable to failure — rework, waste, and corrections that trace back significantly to information and data failures.
5–10%
of project costs go to direct rework
The Construction Industry Institute estimates that 5 to 10 percent of project costs are consumed by direct rework — a significant portion of which originates from information problems.
9%
always incorporate data into decisions
Only 9% of construction professionals always use project data in decision making. 64% do so sometimes, rarely, or never — defaulting to gut feeling under time pressure.
Top 3
poor data as a cause of project overruns
KPMG identifies poor data and information management as one of the top three causes of cost and schedule overruns on construction projects — alongside planning failures and supply chain issues.
77%
of megaprojects delivered 40%+ late
McKinsey found that the vast majority of large-scale projects significantly overrun on schedule — a problem rooted partly in poor information management and data-driven planning failures.
What it costs youFrom global figures
From global figures
to your own numbers.
The global figures are striking. But the real question is what bad data costs your organisation specifically. Use the model below to estimate your exposure.
How the global figure is calculated
€ 108T2025 global GDP
× 13%
€ 14.04T2025 global GDP in construction
× 5%
€ 702BCost of rework
× 14%
€ 98.28BAvoidable rework from bad data
T = Trillion (1012), B = Billion (109).
Estimate your own exposure
Based on the FMI / Autodesk methodology for European construction firms.
Annual revenue (click to edit)
Annual rework cost (5% of revenue)
€50,000.00
Avoidable rework (14%)
€7,000.00
In BIM workflows specificallyWhere bad data enters
Where bad data enters
your models — and spreads.
The industry-wide figures capture the full picture. But in BIM-driven workflows, the problem has a specific origin: data errors that start in Revit and propagate downstream through every system that depends on them.
Missing or incorrect classification
Elements without correct classification codes cannot be reliably used for procurement, cost estimation, or handover. Every system downstream makes assumptions — and those assumptions diverge. Inconsistent parameter values
The same element described differently by architect, structural engineer, and contractor. Three versions of the truth mean no version is trusted — and manual reconciliation fills the gap. Acceptance without verification
Models are approved under schedule pressure without structural data checks. Errors are officially signed off and inherit the authority of a delivered model. IFC export as the point of no return
Once a flawed model becomes an IFC file, every downstream system — ERP, supply chain, construction OS, digital twin — treats its errors as ground truth. There is no automatic recovery. Frequently asked questionsThe questions decision makers
The questions decision makers
actually ask.
Direct answers to the most common questions about data quality costs in construction — optimised for clarity and precision.
What does bad data cost the construction industry?
+
What percentage of construction data is bad?
+
Research by FMI and Autodesk across 1,115 European construction professionals found that 40% of the average organisation's data is bad — defined as inaccurate, incomplete, inconsistent, or untimely. Additionally, 95% of all construction data captured is never used operationally, meaning it delivers no value to the processes or systems that depend on it.
How does bad data affect construction decision making?
+
The FMI/Autodesk 2021 study found that bad project data leads to poor decisions 41% of the time. Only 9% of construction professionals always incorporate project data into their decision making — 64% do so sometimes, rarely, or never, defaulting instead to gut feeling under time pressure. KPMG identifies poor data as a top-3 cause of project cost and schedule overruns.
What causes bad data in BIM models?
+
The main causes of bad data in BIM workflows are: (1) no structural data governance — rules exist on paper but are never systematically enforced, (2) incomplete or inconsistent checking during the modelling process, (3) model acceptance under time pressure rather than data verification, and (4) IFC export without validation, which propagates errors to all downstream systems as ground truth with no recovery point.
How much time do construction professionals lose to data problems?
+
According to FMI/Autodesk (2021), project managers and field supervisors spend on average 49% of their working time collecting, managing, and analysing project data — much of which is ultimately unusable. McKinsey estimates that 13% of engineer working time is lost simply searching for information that should be immediately accessible. This time produces no output and no value.
How can DAQS reduce the cost of bad data?
+
DAQS addresses the problem structurally — not as a reporting tool but as an intelligence layer embedded in the data pipeline. By validating model data against governance rules and downstream system requirements before it moves, transforming validated data into formats that systems can consume directly, and connecting output to operational environments without manual translation, DAQS removes the primary sources of data failure from the workflow entirely.
Get the full evidence base as a PDF.
The complete source data, methodology, and DAQS perspective on the cost of bad data in construction — formatted for internal sharing and business case development.