BackStartups

Chapter 11 in 2021 · over $2bn spent

Katerra

A construction technology company that spent more than $2bn in little over six years attempting to industrialise building, and filed for bankruptcy in June 2021.

What happened

Katerra set out to do to construction what manufacturing had done to other industries: standardise components, build them in factories, and assemble them on site. The thesis was and remains defensible — construction productivity has famously lagged other sectors for decades.

It raised enormous sums, largely from SoftBank, and grew by acquisition as well as by building factories. In June 2021, having spent more than $2bn in a little over six years, it filed for bankruptcy.

What it ran into is the thing that makes construction resistant to industrialisation in the first place: projects are local, regulation is local, labour is local, and the standardisation that makes a factory efficient is exactly what a specific site does not permit.

The capital structure made the problem terminal rather than merely difficult. A business consuming capital at that rate has no capacity to iterate, because each iteration costs a factory.

Visible at the time

What an investor could have seen

These are things that were observable before the collapse, not hindsight dressed as foresight. Some failures genuinely could not be seen coming; where that is the case, this section says so rather than inventing a warning.

  • Vertical integration across design, manufacturing and construction attempted simultaneously rather than sequentially.
  • Growth by acquisition alongside heavy capital expenditure, before the core model was proven at one site.
  • A thesis about industry-wide inefficiency without evidence that the specific inefficiency was addressable.
  • A burn rate that made small experiments impossible.

For your own diligence

What to do differently

  • A correct diagnosis of an industry problem is not a business. Construction productivity really has lagged; it does not follow that a vertically integrated company can fix it, and the distance between those two statements is where the $2bn went.
  • Watch for simultaneous rather than sequential ambition. Companies attempting to change several parts of a value chain at once have multiplied their execution risk and removed their ability to learn from any single part.
  • Ask whether the burn rate permits iteration. A company that must be right first time is a company with no mechanism for being wrong cheaply, and almost every startup is wrong about something.
  • For an angel, businesses that need billions are structurally poor positions: the dilution across the required rounds leaves an early holder with very little, however well it goes.
The guide that covers thisDeep tech & hardware diligenceOpen the checklist →

Checked, not remembered

Sources

Legal outcomes on these pages move — convictions are appealed, sentences reduced, pardons and commutations granted. Every status here is stated with the date it was accurate to, and every figure is traceable to the reporting linked below.

Katerra: common questions

Was Katerra's thesis wrong?
The diagnosis was sound — construction productivity has lagged other industries for decades. What did not follow was that a single vertically integrated company could fix it, given how local projects, regulation and labour actually are.
Why are capital-intensive businesses bad for angels specifically?
Because of dilution. A company needing several billion dollars will raise many times, and an early position is reduced to a very small fraction long before the model is proven. Model the capital path before assessing the technology — it often settles the question.