How Construction Disrupted Katerra - A ConTech Identity Crisis.

Amit Baria
13 min readJun 30, 2021

Owning a construction company is a lot like owning a venomous pet snake. The minute you take your thumb off its neck, it will bite you. Of contractors both large and small, this analogy rings true and the best I’ve heard.

In recent weeks Katerra has fallen from its Silicon Valley perch sending a reverberating thud across startup land. Only as bankruptcy proceedings have come to reveal, it seems not soon enough.

Katerra was committed to taking on the accelerated housing supply and affordability crisis via:

“transformational change through technology — every process and every product.”

Destined to be the future of construction, selling its story to investors, governments and developers that eagerly bought in. With good reason. Using mass manufacturing coupled with technology, Katerra’s model was positioned to produce magnitudes of efficiency to the slow, inefficient, sometimes obsolete and always costly construction process.

Except Katerra’s critical and compounding mistake emanated from an identity crisis.

Its attempt at being a ConTech, or a construction + technology company, missed the mark in that it forgot to be a construction company first. Instead it bet on the tech, or more so a ‘tech-y’, ‘startup-y’ framework, to help accelerate gaining enough critical mass within construction that it would fade out its deficiencies in favor of tech’s new, better way.

For the past two decades we’ve seen technology successfully blaze that disruptive trail leaving incumbent models initially in denial, then unprepared to catch up. However, construction offers a high degree of complexity which defensively positions it. It can’t simply be forced to adopt or else shutter at the same pace as we’ve seen occur in other industries.

Is Construction Un-Disruptable?

Was there promise with Katerra? On paper, a great model. The job site is an incredibly nuanced, living, breathing organism akin to the interdependencies of the human body. All components must function in simpatico to stay alive and thrive.

It’s highly fragmented in its supply chain, and somewhat disparate in regulation across regions. Throw in limited skilled trade labor pools, varying union mandates and bureaucratic red tape.

The various moving parts are management, resource and planning intensive. Equating too slow, expensive and riddled with risk from the project owner, to the prime contractor, down to subcontractors and vendors.

Back to our opening analogy…inviting in technological change can be viewed as attempting to re-grip your thumb, creating a window of opportunity for those fangs to sink in. It’s no surprise for many contractors, the thought of change causes abrupt pause. If it’s not broke, don’t fix it.

Katerra’s model, simplified, was to build off site, transport and assemble on site. This requires less manpower, reduces opportunity for costly mistakes and advantageously enables control of the supply chain from design to build. Technology and vertical integration of the construction process makes for a good model if done by first mitigating the common pitfalls that normally cause a contractor to go bust.

If it’s a good model, how do you standardize to scale a highly specialized industry?

That’s easy. You don’t.

At the helm of Katerra were tech leaders lacking the intricate industry knowledge required to operate true to what it was… a construction company.

Traditionally vertical integration tends to benefit the integrator of the value chain, up and downstream, through cost reduction and quality control. Primarily flowing to their bottom line versus being passed on to the end consumer.

In some respects, technology has spent decades shifting control through dismantling vertical integration in an effort to create transparency and efficiency. Its democratization effects rewarding consumers in the form of economic benefits and optionality.

End product construction (i.e. buildings) are highly bespoke. The complexity, resources and risk are far too great for most developers to undertake directly. (The exception are builder-developers which start out as contractors.) Katerra’s model allowed for a bespoke approach to building where end product developers benefited from the integration sharing in the achieved economies of scale — cost and time. Otherwise, why bother with modular prefab?

Sounds great. What’s the issue?

Vertical integration works best with scale. Scale is achieved best through standardization and automation.

For example, vertical integration for Tesla works much differently than for Katerra, where the former is able to standardize nationally and for the latter it’s near impossible. Tesla needs to abide by Federal Motor Vehicle Safety Standards and Regulations set out by the USDOT. Each car off the line mirrors the previous, is deemed road ready and safe to travel the country. It’s up to the driver to properly operate the vehicle and abide by each state’s individual laws.

In contrast, while project specific, Katerra must abide by multiple federal authorities including the EPA, OSHA and FCR to name a few. Complicating matters are the individual state and municipality regulations along with building codes to adhere to. 50 states. 50 sets of regulations.

Limited standardization in construction is possible. We’ve seen it with prefab and within specialized trades. Katerra’s ambitions took for granted what’s taken many a contractor down, the pursuit of growth over profit. Carillion and Astaldi come to mind. Other foreign contractors have similarly fallen victim to the unknowns, entering the US market using the same approach that has worked for them back home, only on unfamiliar, local levels.

Growth is good…until it become growth by any means necessary.

Getting Fat > Getting Slaughtered

When one small domino falls, the impact is great enough to cause another domino 1.5 times its size to drop. The momentum accelerated faster than Katerra could keep up, which ultimately leveled it. Whether cash from investors or in the form of acquisitions, Katerra was unable to practice portion control. Their growth plan was inorganic and financial resources fundamentally couldn’t support it.

This wasn’t a tech problem. This was a tech industry problem. A model where profits don’t come first. Growth does.

Buying growth and inaccurate reporting.

Contractors buy growth by buying markets. Often in the form of acquiring outfits entrenched within a region or segment of their value chain. It’s a method for contractors to boost revenue and market positioning by expanding beyond their backyards into new geographical territories. The pursuit of buying markets provides access to larger projects, regional teams and infrastructure. This presents a quick means to scale growth versus the organic alternative. One which may be more capital intensive and comes with a steeper learning curve in attempting to crack a local market as an outsider.

Similarly, contractors buy growth by buying work. Or, low bidding jobs to secure work. Aware little to no profit margin is built in. Another treacherous expansion effort in an already risk laden business.

(Note: In some cases it’s justified to keep staff working rather than lose talent to a competitor as many did during Covid.)

As with any investment portfolio, the multiple on winners is expected to make up for the losers. Just as a contractor will rely on a mix of existing and future backlog of profitable projects to make up for the duds. This requires a healthy and diverse order book to fuel growth.

Katerra on the other hand used this strategy in an attempt to grow without booking profits. A newly won job carried the previous losers, and this works…until it doesn’t.

Eventually working capital runs dry. That is, unless Softbank’s Vision Fund is influencing that growth model with follow on funding. After its Series D round, the acquisition spree became a priority, where growth then overshadowed the initial vision.

Cracks in the Foundation

What happens when a business (or contractor) deviates from plan? Stakeholders get nervous. At least they should. There’s been plenty of finger pointing yet the failure to identify and remedy poor practices is simply part of standard operating procedure.

Where was the oversight?

Management / Board

Michael Marks, the praised “Hero of Manufacturing” and former CEO of the year. A Silicon Valley and Private Equity legend. Perhaps…in those fields. Here stems the identity crisis…at the top. Marks instilled a tech industry model on to what’s primarily and most importantly a construction company. Second a manufacturer / tech company. The nuances of construction at this magnitude can only be successfully managed by an industry veteran. Or a tech CEO with several construction right hands. While astute and accomplished, management was filled with PE, tech and manufacturing executives. Katerra ignored its fabric as a contractor.

Why is this?

Construction and Technology, together, are still figuring one another out. Construction corporates are getting better at adopting tech, however there are 700,000 SMB contractors which amass the industry.

Tech believes there is a better way, and more often than not in the past decade it’s been proven that case. Current construction executives paved the path, creating THE way. Construction executives appointed at the most senior levels of Katerra may have identified the trouble signs, insisted on creating guard rails (as they should) thereby stalling the momentum Katerra’s CEO promised and investors pushed for. As for the board, it was composed of renowned tech leaders, real estate investors and manufacturing experts.

Where was the executive with mud on their boots?


Perhaps more head scratching was the absence of a construction CFO. A CFO intimately familiar with the importance of bonding capacity, WIP schedules, revenue recognition, job costing, proper estimating, profit fade signals and every other intricate detail associated with construction accounting methodology. Someone that may recognize signaling of front loaded contracts, long term cash flow constraints created by discounting legacy customers or installing max price guarantees. Someone that can flag improper revenue recognition.

Many of these items can be identified from reconciling the fundamental accounting project management report called the work in progress schedule, or the WIP. This tool displays a firm’s projects in progress, backlog, estimated profit, earnings, completion percentage and more. Think of it as an EKG monitor for a construction firm or a lens lens into profitability for each job.

One would think the annual audit would uncover abnormalities. Signs of financial misrepresentation such the front loading of contracts by the renovations division or poor bidding practices company wide.

Except for that, shockingly, the WIP is not required nor standardized across construction firms and does not fall under common accounting practice, GAAP or otherwise. It’s unaudited and as of now, un-auditable. The information reported is internally provided. Most critical, it flows into the audited financials. In practice auditors are limited to working with information reported within the WIP, with no real way to reconcile.

Venture Capital

Without a doubt, the construction industry displays characteristics that make it ripe for venture investing. A TAM exceeding $1.2T in the US, $12T worldwide, analog processes, obsolete and sluggish legacy tech. The fact that the industry will never disappear. It’s certain, we’ll always need to build and rebuild.

Had the WeWork implosion not occurred with SoftBank as majority lead, perhaps less scrutiny would be occurring here. Another of their Unicorns in only 4 years falls and it begs the question…where was the underwriting?

Venture biases at play? A cognitive bias is a systematic error in thinking, based upon input generalizations and rational deviations from the norm, forming a subjective reality leading to error in judgement or faulty decision making.

Pattern matching and clustering seem evident. Repeat CEO, Silicon Valley guru, KKR partner and HBS alum. Marks built Flextronics into the largest tech manufacturing company in the world during his tenure. This would all surely translate into phenomenal leadership and vision for Katerra.

Marks was a pioneer of the ‘flat’ management style the tech industry is now known for. He was no micromanager and gave executives a long leash in financial and creative decisions.

He’s quoted saying during his Flextronics days,

“I’ve surrounded myself with people who are bright and enthusiastic and don’t want a lot of direction. We grew at 60 percent a year for six or seven years. If you don’t have this kind of organization, you can’t grow like that…. I set expectations and judge people by results. By definition, if you give people a lot of rope, some will hang themselves. So I occasionally fire people”

And his COO, Mike McNamara stated,

“Marks lets you run your own business as long as you produce results. He doesn’t believe in complex structures, reports or endless reviews. He allows us to innovate.”

A style that’s been quite successful within tech and appreciated by management and employees. An industry that is known for failing your way to success through constant iteration.

Construction is a different animal. Where margins can be paper thin, there is little room for error. The constant feedback loop of meetings, reviews and reports are simply unavoidable. It’s the only way to keep your thumb on its neck. Not surprising Katerra’s renovation division wasn’t caught sooner.

Planning fallacy is evident at every raise. Construction is slow moving and timelines are often not met. Contingency time is usually added to provide an extra buffer. Revolutionizing the construction industry with technology is no exception. There was no built in contingency in growing to hit their 2020 profitability target.

Circling back to WeWork, how about anchoring? Rather, where was it? Painful experiences are known to create strong anchors. SoftBank posted a $6.6B loss on WeWork in 2020. That would create a strong anchor for most investors, leading to additional scrutiny in underwriting, increased measures to identify red flags and prevent repeat mistakes of the past. Instead, quite the opposite as blinders went up neglecting signals that the model was in severe jeopardy.

Throwing good money after dead money isn’t a sound venture strategy. Only perhaps the Vision Fund is too big to care. At $100BN+, Vision Fund can spread sizable bets across the table.

When Unicorns like Coupang helped generate $45B this first quarter it’s easy to dust off a mere $2.5B bet gone bad. It’s the venture model, where 10% of your wins generate outsized returns. Multiples which wipe out losses and generate astonishing performance fees.


Considering 70% of revenue in the surety business is derived from bonding construction projects, it’s no question they’re intimately familiar with the contractor business model.

Underwriting consists of evaluating the contractors character, capability and capital to complete projects. As qualitative as it is quantitative, its credit underwriting model can be more comprehensive compared to bank debt and certainly equity investors. It allows bonding companies to be perhaps the best predictor of a potential contractor implosion.

Was there a failure to consider the audited financial reports, deviations from business plan, recurring and accumulated losses, lofty growth plan and critical WIP signals? Bonding companies too, don’t get paid for intellectual laziness.

Eventually this became the dagger that did Katerra in. Unless the sureties had the benefit of the parental guarantee from SoftBank or the fund’s uncalled capital, the financial woes, bidding strategies, overruns and aggressive growth plan would have caused an underwriter pause in 2018.

In this instance it’s hard not to believe there wasn’t pressure to keep bonding capacity open to where the sureties loosened underwriting standards. Perhaps influenced by the deep pockets of SoftBank as a permanent backstop?

While SoftBank did up their ante for $200MM last fall, the losses called for a more significant capital injection, collateral requirements, retention agreements or other liquidity control measures necessary to continue surety credit support. And that’s after the Renovations division was flagged for improper revenue recognition, then subsequently probed by the SEC. Add in the uncertainty that Covid provided where many bonding companies tightened underwriting standards. You get the picture.

It was only after Greensill, another SoftBank backed lender went south earlier this year, that reported surety collateral requirements spiked (this suggests SoftBank’s parental indemnity was not provided). No surprise there. Project bond mandates began to increase as Kateera’s liquidity problems became public. Bond demand overtook now capped surety supply.

Without ability to meet bond mandates, Katerra was now disqualified from bidding additional work to ‘grow’. Softbank also cut further support squeezing liquidity (unlear as to who was first, the VC or the Surety).

The ride’s over. The bonding company is a contractor’s partner in growth, particularly when bonds are mandated to bid new work. Again, Katerra ignored it was a construction company first.

Transformed, not disrupted.

Katerra had a lot going for it in addressing a commendable societal problem. It began with the vision of solving the affordable housing issues Americans face. It was on that path to set the new standard of constructing residential buildings, serving as the exemplary model in simplification and cost reduction of the process using technology.

Over $2B invested later, it still failed. Although not at the fault of the technology. Nor the vision.

The technology is here to stay. The modular building market has seen modest, steady growth and is expected to comprise 30% of the North American market by 2028 totaling $115B up from today’s $72B.

Katerra’s biggest asset ended up being an equally large constraint — capital.

Not every industry is subject to the same blueprint of disruption — ‘move fast and break things’. Not every industry problem is solved by deploying hefty sums of venture investment into it.

In construction, if you break something, it may break many more things and fixing it can be quite expensive. Pressure from venture investment to generate returns for shareholders eventually overtakes the slow, steady, consistent growth to profitability contractors find success in.

Katerra’s identity crisis was in taking on the persona of a Silicon Valley startup when its purpose was to be a Contractor. The torque behind the startup playbook and taking on too much investment, too fast ripped Katerra’s engine right off its chassis. Th unfortunate conclusion that technology in fact would not outwork and overcome the common contractor pitfalls that have taken down other giants.

So, can construction be disrupted?

Disruption comes with a cost far too great. It’s transformation instead that is key. Construction isn’t easy. Tech can certainly help it become better, and it is. ConTech startups should continue tackling construction’s problems despite Katerra’s optics. Arguments against R&D within construction are mistake. Change is imminent and progress is being made by the remaining ConTech universe that don’t benefit from the cash-ay Katerra touted.

Tech alone is not the answer. Construction alone is not the answer. ConTech isn’t dead, it’s reached a defining moment. Tech needs to know they shouldn’t try to do it alone. As Mark Andreeson put it [we] “need to want these things more than they want to prevent these things.”

Construction should accept that tech will happen, and great things will follow. Tech needs from Construction a few lessons in how to charm the snake.

Slowly and with care. Disrupting it risks falling victim to a venomous bite. For Katerra, one that was lethal.