Katerra raised more than two billion dollars to industrialize construction. It went bankrupt in 2021. The widely accepted lessons focus on overreach, vertical integration, and SoftBank’s misjudgments.
Katerra was the loudest failure but not an isolated one. Foundamental’s State of the Project Economy 2026 documents that modular construction funding peaked at $298 million in 2021 before falling sharply, as unit economics proved harder to crack than the wave of capital had assumed. The same pattern is visible in Europe, where well-funded modular ventures have closed, scaled back, or struggled to find sustainable unit economics despite favorable demand fundamentals.
The less discussed lesson is more useful for the rest of the industry than the headline failures: getting the factory right matters less than getting the information right before manufacturing begins.
That lesson now applies broadly. Modular construction is a $111 billion global market in 2025, with Europe holding the largest regional share. Skilled labor shortages across North America and Europe are pushing more developers and contractors toward factory-built delivery. But the preconstruction process feeding most of these projects is still designed for stick-built work. That mismatch is the most important unmanaged constraint on industrialized construction’s promise.
Why preconstruction matters more in modular work
In conventional construction, decisions can be corrected on site. A misread specification becomes a change order. A misunderstood contractual obligation becomes a claim. The cost is real but usually absorbable.
Industrialized construction operates under different constraints. Design information must be coordinated earlier. Procurement commitments happen sooner. Manufacturing workflows depend on stable requirements. By the time a module leaves the factory, the decisions that shaped it are mostly locked. An overlooked specification or contractual ambiguity does not just create rework. It cascades through fabrication, logistics, and installation, often with no on-site path to recovery.
From document management to decision management
Most preconstruction work today is organized around documents. Teams review tender packages, cross-reference specifications, and rely on experienced estimators to catch what matters. This worked when project documentation could fit in someone’s head. It no longer can.
This calls for reorientation more than new tooling. Preconstruction has to become a decision discipline rather than a document discipline. The question is no longer whether the documents have been reviewed. It is whether the organization knows what it is committing to, where the risks sit, and on what basis they were accepted.
The firms that outperform their competitors are often those that answer fundamental questions earlier and more confidently:
- What are we actually committing to?
- Where are the contractual risks?
- Which requirements drive cost and schedule?
- What assumptions are embedded in our bid?
- What obligations will affect project delivery?
- Are we pursuing the right opportunity at the right terms?
These questions sit at the intersection of commercial, technical, operational, and contractual decision-making. They determine project outcomes long before project execution begins. When commitments lock earlier, the decisions behind those commitments need to be more deliberate, more traceable, and more defensible.
Two capabilities, not one
Most conversations about preconstruction collapse into a single goal: make better decisions. That framing misses half the problem. There are two distinct capabilities at play, and they require different investments.
The first is decision quality. Did we identify the right risks, price them correctly, accept the right obligations, and turn down the wrong opportunities? This is what experienced estimators and bid managers have always been measured on. It is the capability the industry knows how to recognize, even if it does not always know how to scale.
The second is decision traceability. Can we reconstruct, after the fact, what we knew when we made a commitment, what assumptions we relied on, and on what basis we accepted a given risk? This is the capability that determines whether a project organization can defend its decisions during execution, in claims, or in the next bid cycle.
These are not the same thing, and they do not scale the same way. Decision quality improves with judgment, experience, and pattern recognition. Decision traceability improves with structure, instrumentation, and discipline in how information is captured and surfaced. An organization can have high decision quality and low traceability, which leaves it strong individually but fragile institutionally. The reverse is more common in larger contractors: well-instrumented processes that still produce uneven decisions because the underlying judgment is not consistently informed.
Industrialized construction stresses both. Lock-in happens earlier, so decisions must be sharper. Errors propagate through the factory, so the path from commitment to consequence must be reconstructable. The capability gap most modular ventures have hit is not one or the other. It is the absence of either.
Where AI actually changes the work
AI is generating significant noise in construction. Much of it is misplaced. Automating routine tasks is useful but limited. The more meaningful shift comes from a narrower application: helping people extract decision-grade understanding from unstructured project information, and creating the traceable record that lets decisions be defended later.
Construction has always been an information business operating under a project structure. Contracts, specifications, drawings, and standards interact in ways that resist surface reading. Companies working in this space, Volve among them, focus on making those interactions visible. The objective is not to replace judgment. It is to ensure that judgment is applied to the right things, with the right context, before commitments become irreversible.
What this means for executives
The next phase of industrialized construction will not be won by whoever builds the most efficient factory. Several well-capitalized companies have already proven that factory efficiency without preconstruction discipline produces expensive lessons, not durable businesses. It will be won by organizations that treat preconstruction as a strategic capability and invest in decision quality and decision traceability as distinct things. The most expensive errors rarely originate in the factory. They originate in what was, or was not, understood before the first module was ordered, and in whether anyone can reconstruct why. Organizations that recognize this early have time to build both capabilities. Those that do not will keep paying for the same kinds of mistakes Katerra did, on smaller and quieter scales.
