Katerra wanted to transform construction. It burnt $2B and shut down | Hacker News Reader

Katerra wanted to transform construction. It burnt $2B and shut down

businessbrief.xyz

43 pointsbluemoon201032 comments

The best article I've read about Katerra is still from Construction Physics, it goes into much more technical detail. [1]

[1] https://constructionphysics.substack.com/p/another-day-in-ka...

Thanks. This is a much more interesting read. It doesn’t seem like an obviously bad idea. Perhaps some amount of being too ambitious, some amount of bad luck (WeWork, Greensil), some amount of mismanagement (slow execution, acquisition issues) - but nothing that says the idea was fundamentally bad.

A better, more detailed writeup is here: https://constructionphysics.substack.com/p/another-day-in-ka...

I've been very interested in the Katerra story for a few reasons:

- I'm a real estate investor

- I have a logistics background

- The founder (or some other exec, can't remember) came to speak to my MBA class a few years back, and we toured their factory

What Katerra was trying to do is highly compelling, and I have to believe it'll eventually get done.

In terms of how it gets done, two paths initially come to mind:

1) Amazon acquires to leverage the logistics-at-scale approach and funds long enough to find P/M fit.

2) A new startup (or spin-off) focuses on one of the products (I'd choose software first) and nails P/M fit. By strategically making the first product one that removes blockers to pave the way and jump-starts change in the rest of the industry they can expand into adjacent products. Maybe there are existing companies like Houzz that could incubate one of these projects.

IMO solving this and combining it with ESG is world-changing to the extent of Tesla--affordable, sustainable housing for everyone. It could even be a status thing where the wealthy are willing to pay a bit more and for every build a home gets built for someone in need. If I were back in my 20's with no kids/mortgage and looking for a new project, I would tackle this.

I was under the impression that the “housing problem” is basically land prices anywhere near where everyone wants to be, and a zoning/regulation/NIMBY thing, not actual construction costs. Am I wrong?

Yes, you are correct. The housing prices have been steadily climbing over the years, and that’s not because companies are getting worse and worse at building houses, and it hasn’t been because of rising material costs (the rises that currently are are all much shorter term and due to corona). Essentially every policy or local decision that could help to provide more affordable housing has also been a direct threat to the real estate investments of the politicians who have shut it down. It’s the biggest generational theft ever seen, where the older generation is deliberately denying the entering generations cheap access to zoning in order to inflate housing prices and extort as much money as humanly possible from the youth.

Children and the very old can’t work at all. So the youth in the middle have to do the work if they ever expect to make it out of childhood and into old age. how could it be any other way ?

Just because youth must do some small amount of work to make it to old age doesn’t imply that they aren’t being massively extorted.

Youth must do all the work. Who else would do it ?

Not all work needs to be done.

Almost all work doesn’t need to be done. We are talking about the work that gets done.

Reading about Katerra here and elsewhere brought me to a naïve question not easily answered by Google:

How often are building blueprints re-used?

E.g. I have a company and we need to expand and want our own building. Do we really need to hire an architect to design something new? Surely someone has already designed a building in the last 5 years that would suit our purpose. Just buy those blueprints and make it again (with some slight adjustment for local codes, geography, etc.). Perhaps this is done frequently and simply not very noticeable?

(I understand the same blueprints may be reused for multiple buildings in one complex, and more often in the public sector).

There are house plans you can buy, but I think they require a pretty flat site? It probably works better on a larger lot too.

I wonder what licence or ownership the building owner has over the architectural drawings & blue prints.

I imagine it wouldn’t be in an architects interest to have the plans sold multiple times. Tho on the flip side, it’s unlikely the plans would be used without changes, minor or major, which are likely best made by the original architectural firm. Which could become a source of additional business.

It's common for the architect to stipulate in the contract that the plans are for the specific project (down to the location) and cannot be used for any other project.

From what I've read/heard, reusing blueprints to a house is one of the more economical ways of building a new house. You start with a pre-made blueprint that has the necessary approvals, and do modifications to it as you please. Usually this should result in less bureaucracy to get the house built than designing from scratch.

I think it is done frequently in commercial development, just not as a product that would be sold directly. The whole point of hiring an architect is so that you don't have to build that team within your own company. Even if you bought a set of blueprints you still need expertise to check, adjust, sign-off and take responsibility for that. And you have to get all that through corporate management and politics. All for something that is probably not your core business. You might as well hire an architect.

And I doubt that the plans are even a large proportion of the price. There is communication, project management, scheduling, keeping aligned with all the other specialists, insurance, legal fees and just normal overheads. And when you want to build another building of a similar design the architect can reduce their fees to take advantage of reusing existing content. The architect becomes the go-to for a particular kind of development and can win bids based on experience.

If you can do it in-house the designs become proprietary and won't be shared externally.

Many many years ago (think about late '80's, early 90's) I was somehow involved in a division of the building company I was working forat the time (in Italy) dedicated to pre-fabricated buildings (not your typical solo villa, rather multi-apartments blocks).

The head of the division was a very experienced and extremely smart technician, with some 20+ years expereience specific to pre-fabrication.

The base ideas were three:

1. have a small number of pre-made, tested, industrialized/industrializable designs/blueprints, with a somewhat "modular" approach

2. build these faster and cheaper

3. profit

At the end of the experience (several years/projects built) the results were that there was no way to re-use the same blueprints on different projects, due to a number of different reasons, local rules, size of the land lot, preferences (either technical or purely aesthetic of the client), whatever, each blueprint had to be modified/adapted/changed and at the end it took as much time and money as starting from scratch a new design.

Then it came out (again, in practice) that building pre-fabricated (and then transported and assembled) at the same (high) quality level as built on spot costed roughly the same.

It was only much faster, but I believe this mainly because you built panels/structures/etc. with highly specialized people, inside a perfectly tuned pre-fabrication area, protected from bad weather, with all the materials readily available, etc., and because of the much more accurate planning possible.

No profit.

This does not generally happen, at least full copy and paste. What usually gets used are the “details” which are the parts of the blue print modern terminology is Construction Drawings or CDs that show very zoomed in specifics which will be the same in many cases regardless of the global structure.

There are major incentives and reasons why full copy and paste doesn’t happen more often some are cultural and some are technical.

Culturally because construction projects tend to be very expensive architects sell the client by “building the client’s vision”. So the incentive is to keep making unique buildings every time. In this way architects earn their keep and along the way if they can make their current client happy it may turn into a bigger future project.

Technically speaking a new location introduces new variables which the old building design may not meet. For example there may be new risks of earthquakes, hurricanes, tornados, blizzards, high heat/moisture, poor soils (this is the big one). Additionally a new location may require new construction materials or ways of building/methods. Any one of these could lead to large design changes. When the risks might not change our understanding of them might so we may need a new design due to new building code updates. Also building codes and requirements are controlled at the local level so they may vary by legal jurisdiction so even though all else is equal the original design does not meet the code governing the new location.

Interestingly there was an open source house project a couple of years ago with the ambition of over designing it so that you could build it in any part of the US. I personally love the idea however I am very skeptical due to the way construction liability works. So say there was a collapse who does the owner sue? Is it the designers all 100 of the people who contributed to the project? What if some aren’t engineers but hobbyists? What if they don’t have engineering insurance? Do they sue the contractor who built it? As you can see it gets quite messy. Construction being an old and mature industry has a lot of regulations about legal responsibility which IMO don’t play well with the modern day tech ethos of “just build”.

This is really interesting, I’d love to read more. One thing that’s interesting to me is that when I search HN for Katerra, I see plenty of articles posted that verify the OP story, but none of them have any comments. I guess it’s not really a tech company so even with huge valuations and bailouts it didn’t really attract interest in discussion at HN.

My company is similar. One of the fastest companies to “Unicorn” status ever with rounds led by top firms, but garners nearly zero discussion on HN.

Could you tell what's your company? I don't believe in hype and so keen on great companies flying under the radar.

What went wrong? It doesn't say.

Looks like it’s the first part of a multi part series

I read that as the first of a multi-part series of SoftBank investment failures, not a multi-part of this specific one.

The company made big promises and had to burn through the funding to try to meet them without succeeding.

Another “revolutionary” construction company to add to the ash heap of failures.

"This blue-collar profession turns out to be way more complicated than we expected" seems to be a common theme.

“How hard could it be?”

Computers solve a lot of problems, as long as those problems are digital by nature. So naturally taking over an amateur mathematician who made his living as a computor, that was an obvious one. And we've gone through a cultural transition to solving all our problems digitally: using the internet to get all of our information, listening to digital music that is in a way a nostalgia of what the real sound was like, even trying to get machines to do 100% of our driving.

Blue collar? No, it doesn't work digitally, a huge amount of a blue collar worker's knowledge is not in his skull, but in his hands. Computers are remarkably bad at solving analog problems, just witness how terrible online dating is, it sucks. Or online classes, a totally inferior substitute.

Another Soft Bank investment so that was pretty certain.

This doesn't necessarily mean that every company that will try to transform the construction industry will fail. Vertical integration definitely makes sense to eliminate inefficiencies. Perhaps, a new company could better execute vertical mergers and acquisitions and succeed.

Krattera.

The construction industry is being revolutionised quickly.

"Almost every advancement in construction is small enough for a man to carry" - https://austinvernon.site/blog/construction.html

So if Katerra wanted revolutionary construction that jumps even quicker than this ongoing well documented advancement, Elon Musk style from the ground up, they need a reason why their product would work contrary to what anyone can see. They never gave one, so I don't see why it would have worked.

Prefab needs a prefab factory in every city. And think of how limiting in design that would be.

Kick ass screws need one factory in the world. And the fact people get militant about things like Square-Drive screws means screws will continue to be pushed evolutionarily. Who knows if the Phillips head can be upended but we are at a time in history that we can find out.

Reply on news.ycombinator.com