12. The Cost of Disconnected Project Data
Disconnected project data is information held across separate Common Data Environments (CDEs) that do not exchange with each other, so the same document has to be moved by hand from one to the next. The cost of disconnected project data is what that manual movement produces: labour spent transferring and re-keying, errors that surface later as rework, decisions that wait on documents already issued somewhere else, and exposure when a claim arrives and the record cannot be reconstructed.
Very few organisations have a figure for it. The spend is spread across document control salaries, variation orders, extension of time claims and legal fees, and none of those cost codes traces back to the data problem that produced them.
This article covers:
What disconnected project data is and why the cost stays unmeasured
Direct labour spent moving and re-keying documents
Error and rework cost, the largest and least visible layer
Delay cost when a decision waits on a document that already exists
Dispute exposure without a clean, traceable record
A worked cost model and the measures to take before and after
What is disconnected project data?
A project has disconnected data when the client platform, the designer platform and the supply chain platforms hold overlapping information and no automated link between them. Documents cross those boundaries by download, rename, re-key and re-upload.
Eleven environments is a normal count. Deloitte and Autodesk found in their 2025 State of Digital Adoption research that construction businesses across Asia Pacific use a median of 11 separate data environments, and estimated that streamlining them could return 10.5 hours per week.
The industry-level totals are large enough that boards discount them. Autodesk and FMI Corporation, surveying more than 3,900 construction professionals for their 2021 report Harnessing the Data Advantage in Construction, put the global cost of bad data at an estimated US$1.85 trillion in 2020, attributed 14 per cent of all rework globally to it, worth US$88.69 billion, and calculated that a contractor with US$1 billion in revenue could be carrying up to US$165 million a year. A project board will not approve a case built on a trillion dollar figure. It needs four layers it can recognise on its own job.
Layer one: labour spent moving and re-keying data
Direct labour is the easiest layer to count and usually the smallest of the four. A case built on this layer alone tends to look underwhelming.
PlanGrid and FMI, in their 2018 Construction Disconnected study, found that construction professionals spend 35 per cent of their working time on non-productive activity: 5.5 hours a week searching for project data, 4.7 hours resolving conflicts and 3.9 hours dealing with mistakes and rework. The same study put wasted labour across the US construction industry at US$177.5 billion a year.
Counting the layer takes three inputs: documents moved each week, average handling time per document, and a fully burdened hourly rate. Handling time runs longer than people estimate, because it covers locating the current revision, downloading it, renaming it, re-keying metadata on the far side, uploading and confirming it landed.
Layer two: error and rework
Rework is the largest of the four layers and the hardest to attribute. The Construction Industry Institute puts rework at 5 to 12 per cent of total project cost. On a A$400 million package that is between A$20 million and A$48 million.
Only part of that traces to data movement, and the coordination evidence is specific about the share. Dodge Construction Network's September 2024 SmartMarket Brief, Not By Design: The True Cost of Poor Collaboration, found that 98 per cent of contractors in the US and Canada experienced serious quality issues in the past three years, and that coordination issues cause an average 9 per cent budget increase and erode 10 per cent of annual profit margin.
Each manual transfer is an opportunity for the wrong revision to move, a suitability code to be dropped, or a metadata field to be entered differently on the far side. None of that shows up at the point of transfer. The error surfaces on site some weeks later and gets priced as a variation.
Layer three: delay while decisions wait
Delay cost arises when a decision waits on a document that already exists in another environment. A hold point stays closed, a crew stands down, or a submission slips while the document that would release it sits on a platform the deciding party cannot see.
Dodge Construction Network's 2024 brief found that only 11 per cent of field personnel always have the information they need. Most organisations know the value of a project day from their extension of time claims. Size this layer by multiplying that daily figure by the number of decisions each month that stall on document access.
Layer four: dispute exposure
Dispute cost is driven less by the existence of a claim than by the work needed to answer it. Establishing what was issued, to whom, at what suitability and on which date takes weeks of reconstruction when the platforms involved never shared a version history.
Arcadis, in its Global Construction Disputes Report series covering 2022 to 2024, put the average construction dispute value at US$52.6 million globally, with sums in dispute trending toward 32.3 per cent of project CAPEX. Hwang and colleagues found in 2014 that every A$1 invested in documentation quality management returns A$7.40 across infrastructure projects, and disputes are where a large share of that return appears.
The exposure continues past handover. NIST, in its 2004 study of inadequate interoperability in the US capital facilities industry, estimated US$15.8 billion a year in avoidable cost, two thirds of it borne by owners and operators during operations and maintenance.
Building a model for the cost of disconnected project data
The table below sets out a structure and an arithmetic path. The middle column holds illustrative placeholder figures chosen to show how the calculation runs. They are not Utopia Digital research and they are not benchmarks. Replace every one of them with a measured input from your own project before the paper goes to a board.
The total excludes delay and dispute exposure, so it should be presented as a floor rather than an estimate. Its usefulness comes from every input tracing to something a reviewer can verify.
The other side of the ledger needs the same treatment. Building an integration in house carries developer time, ongoing API maintenance and the cost of every vendor schema change, which is set out in why custom CDE integration development costs more than you think. Subscription costs are published on the product pricing page, and the full landscape of options is compared in the CDE integration comparison guide.
Australian figures carry a local board further than global ones. Oxford Economics and the Australian Constructors Association put lost productivity in Australian construction at A$62 billion a year. Infrastructure Australia's 2025 assessment describes a A$242 billion five year major public infrastructure pipeline against a shortfall of 141,000 workers.
Measuring the cost of disconnected project data before and after
A business case that cannot be proven afterwards will not be approved a second time. Take a baseline before anything changes, then repeat the same measures at 90 days and at 12 months.
Weekly transfer volume and handling hours. Count documents crossing each platform boundary and time a sample of transfers end to end.
Metadata error rate. Audit a sample in the destination CDE for wrong revisions, missing suitability codes and mistyped fields.
Currency of the receiving environment. Record the age of the newest revision in the receiving CDE against the issuing one.
Decisions stalled on document access. Log the count each month and the average wait, from the RFI and hold point registers.
Evidence assembly time. Record the hours spent gathering the record for the most recent claim or audit.
Where your organisation sits across those measures today is what a digital maturity assessment is built to surface.
What does disconnected project data actually cost a project?
It costs across four layers: labour spent moving and re-keying documents, rework caused by wrong revisions and lost metadata, delay while decisions wait on documents held elsewhere, and dispute exposure when the record cannot be reconstructed quickly. Direct labour is the simplest to count and usually the smallest. Rework is the largest and the hardest to attribute back to its cause.
How do I calculate the ROI of CDE integration?
Baseline the weekly hours spent transferring documents between environments and price them at a fully burdened rate. Add a measured rework rate for transfers arriving with an error. Compare the annual figure against the subscription cost of automating the transfers. Keep delay and dispute exposure out of the headline number and record them separately as unquantified upside.
Is rework really caused by data problems?
Partly, and the share is measurable on your own project. The Construction Industry Institute puts rework at 5 to 12 per cent of total project cost, and Dodge Construction Network found in 2024 that coordination issues alone drive an average 9 per cent budget increase. Sampling your rework register for causes that trace to a wrong revision or missing metadata gives a project-specific figure.
Build the model on measured inputs and keep the source attached to every borrowed figure, so anyone reviewing the case can check it. CDE Sync™ removes the manual transfer layer, with a first sync running in 15 to 30 minutes and no per-seat fees. Get in touch if you want help taking the baseline before the paper is written.