A few months ago I was at an industry event, listening to a panel about digital transformation in construction. The moderator asked a question that should have been easy: "What's the biggest barrier to data sharing on your projects?"
The structural engineer said technology. The architect said training. The client said culture. Then the contractor leaned into the mic and said something that made the room go quiet: "The contract. The contract is designed so that the less I share, the better off I am."
He wasn't being cynical. He was being honest. And he was right.
The adversarial machine
Most construction projects still run on contracts that were designed for an adversarial world. Design-bid-build. Lump sum. Lowest price wins. The designer throws documents over the wall. The contractor prices them. Everyone protects their position.
In that structure, information is leverage, not a shared asset. A contractor who discovers a cheaper construction method has zero incentive to share it — that's their margin. A contractor who spots an error in the design has an incentive to stay quiet and convert it into a variation claim later. A subcontractor who shares detailed pricing data risks being shopped against a competitor offering a lower number.
McKinsey put it directly: fragmentation in construction has led to "hostility and change orders rather than productive and trusted collaboration." That's not a cultural observation. That's a structural one. The hostility is a rational economic response to contracts that reward self-interest.
The trust deficit in numbers
FMI and Autodesk surveyed 2,527 construction professionals worldwide. What they found should alarm everyone in this industry:
- 63% of construction organisations report less than "very high" trust levels
- High-trust organisations enjoy 2–7% higher gross margins than similar-sized firms with average trust
- 43% of high-trust firms feel confident meeting schedules, versus only 21% of average-trust firms
- High-trust organisations save up to $4 million annually from improved deadline compliance alone
So trust literally pays. But the contracts most projects are built on actively erode it.
The $60 million symptom
When trust breaks down, disputes follow. And the numbers are staggering. The average construction dispute in North America is now worth $60.1 million — the highest in 13 years. Claims and disputes add an average of 33.2% to project budgets. Schedule overruns average 66.5%. And 70% of construction lawyers expect dispute volume to increase.
The leading cause of disputes, three years running? Errors and omissions in contract documents. Not site conditions. Not material prices. Bad information.
Here's the cruel irony: the same contracts that discourage data sharing are generating the information gaps that cause the disputes. The system is producing the problem it claims to manage.
Why subcontractors won't open up
If you want to understand the transparency problem at its sharpest, talk to subcontractors. They face the most acute version of the dilemma because they have the least contractual power and the most to lose.
When a sub shares detailed cost breakdowns, they expose their pricing structure to a general contractor who may use that data to negotiate lower prices from a competitor. Research from Brigham Young University confirms this is not paranoia — the fear of bid shopping causes subcontractors to inflate their pre-auction bids and reduces their willingness to participate at all.
Sharing knowledge — key calculations, analysis methods, proprietary approaches — is perceived as giving away competitive advantage. And in an industry where 69% of project owners blame contractors when projects underperform, transparency feels less like collaboration and more like handing someone the evidence to assign blame.
So subcontractors do the rational thing: they share the minimum required, protect their data, and keep their detailed knowledge behind the firewall. The project suffers. But the sub survives.
The "who owns the model" problem
BIM was supposed to fix all of this. A shared digital model. A single source of truth. Everyone working from the same data.
Except nobody can agree on who owns it. Research from the International Bar Association found that a BIM model developed collaboratively cannot be the sole property of any one party — it's jointly owned, and using it without consent of all contributors is technically copyright infringement.
Then there's the liability question. When BIM models contain inaccuracies, who's responsible? The party that modelled the element? The party that coordinated it? The party that relied on it? In a federated model where elements interact and depend on each other, drawing clean liability lines is nearly impossible.
And the insurance industry has not caught up. Professional indemnity policies are designed for adversarial relationships where each party produces independent work. When responsibility lines blur in a shared model, insurers either price conservatively or exclude collaborative BIM work entirely. Professionals are financially penalised for transparency.
IPD tried to fix this
Integrated Project Delivery was explicitly created to break the adversarial cycle. Multi-party agreements. Shared risk and reward pools. Open books. Everyone financially aligned around project success rather than individual position.
And when it works, it works brilliantly. IPD projects consistently report better outcomes on cost, schedule, and quality.
But adoption has been glacial. Because IPD requires trust to function, and it's designed for an industry that lacks trust. Opening your books means exposing your profit margins, your overhead rates, your cost structures. The leap of faith is enormous. And most firms — understandably — aren't willing to make it without proof that the other side will reciprocate.
It's a chicken-and-egg problem. You need trust to collaborate. You need collaboration to build trust. The contract is supposed to bridge the gap, but most contracts widen it instead.
What a neutral data layer changes
I've come to believe the solution isn't better contracts — although those would help. The solution is reducing the stakes of data sharing so that transparency becomes less risky in the first place.
When every discipline's model feeds into a single federated environment — not hosted by any one party, not controlled by any one stakeholder — something shifts. The data becomes project data, not party data. Changes are tracked automatically, so there's no ambiguity about what was shared when. Quantities derive from the model directly, so there's no argument about who counted what.
You can't weaponise information that everyone can see. You can't claim you weren't told about a change when the change is logged, timestamped, and visible to everyone it affects. The information asymmetry that adversarial contracts depend on starts to dissolve.
Does this solve the trust problem entirely? No. Trust is built over time, through repeated interactions, through contracts that reward honesty, through an industry culture that values outcomes over positions. That will take years.
But a neutral, auditable, shared data layer takes away the biggest excuse for opacity: "I can't share because it will be used against me." When the data is shared by default and the audit trail protects everyone equally, the calculus changes. Transparency stops being a risk and starts being a shield.
The $1.8 trillion question
Bad data cost the global construction industry an estimated $1.85 trillion in 2020 — roughly 16% of the industry's total value. Around 95% of construction project data goes unused. Not because it doesn't exist, but because it's locked in silos, protected by contracts, and hoarded by parties who have been taught — by decades of adversarial practice — that sharing is dangerous.
The question isn't whether the industry can afford to be more transparent. It's whether it can afford not to be.
Every dispute that costs $60 million. Every rework cycle driven by outdated information. Every coordination failure caused by a version someone didn't share. Every claim built on an ambiguity that transparency would have prevented. These are the costs of a system that rewards secrecy.
The technology to share data safely exists today. The contracts are starting to catch up. What's missing is the infrastructure that makes transparency the path of least resistance — where sharing data is easier than hiding it, and where every party benefits from the truth being visible.
That's not idealism. That's economics.
Harsh Singh is the co-founder and CEO of Criad, an AI-powered BIM platform that federates construction models into a single live source of truth. If you think your project data should work for the project, not against it, book a call or join the waitlist.