Urban Technology at University of Michigan week 236
How "Chain Drain" holds the AECO industry back and is a space of opportunity for urban technology
Next week I'll be in the Netherlands for the Living Data conference hosted by TU Delft, and the event's focus on "bridging the empirical methods of science and the creative processes of architecture" has sparked some thoughts on a favorite topic of mine: the business model(s) of the built environment. This week’s issue is a brainstorm on how the data is important in this context, the damage done by "chain drain," and what it all means for people interested in urban technology.
💬 Hello! This is the newsletter of the Urban Technology program at University of Michigan, in which we explore the ways that data, connectivity, computation, and automation can be harnessed to nurture and improve urban life. If you’re new here, try this short video of current students describing urban technology in their own words or this 90 second explainer video.
⛓️💥💧 Beware of Chain Drain
Here we go, in the numbered assertion format cribbed from Julian Bleecker:
More data generated by and about cities and buildings promises to unlock more precise understanding of urban challenges, greater creativity in problem solving, and the potential to manage outcomes toward desired targets (lower carbon, healthier air, etc).
That's all good, but is the work of the built environment structured to make the best use of data?
No.
We first need to solve the issue of chain drain so that people who are serious about improving cities can work with greater efficiency.
The acronym AECO stands for Architecture, Engineering, Construction, & (Real Estate and) Operations. AECO represents the full lifecycle of the built environment, from design and construction to ongoing operation and maintenance.
The AEC sector is responsible for 11.9% of GPD, on average, in 146 countries surveyed by World Bank's 2005 International Comparison Program. Add the "O," and it would be even larger.
Buildings and construction are responsible for 37% of carbon emissions according to UNEP.
You would think that if the built environment is such a massive sector financially and has such an impact on climate that it would also be a sector full of innovation.
Reader: it is not! 😭
Historically, AEC (and especially construction) has been slow to adopt new technology, and digital technology in particular.
If you've ever seen a construction site you know this, but if you want a citation for it, McKinsey's got you. They figure a 48-60% productivity boost in construction alone is waiting to be realized through digitization.
It's crazy hard for individual firms to solve this because the vast majority of buildings, public spaces, an infrastructure projects are delivered by temporary syndicates of firms.
Firms that do use emerging technologies are often held back by the need to be "backwards compatible" with the traditional practices of other firms in their business ecosystem, including suppliers and partners.
This logjam is the result of the AECO industry being horizontally organized. Any one project requires collaboration of numerous firms: architects and engineers designing, construction company (with an array of subcontractors) building, and the owner operating.
This horizontal process has historically allowed each segment of the supply chain to optimize (for quality) and compete (for lower cost to client), allowing the owner of the building to shop for the best price at each step and maximize their savings overall.
But the the horizontal complexity of the supply chain also makes it harder for the buildings to be realized that truly deliver on complex performance criteria. At each handoff in the AECO supply chain, intelligence drips out of the process like so many tears:
Handoffs between the firms involved in delivering a building or public space also chop up the feedback loop between design intent and actual lived outcomes.
As the agency of the firms involved is parceled into smaller pieces, it becomes complex or even impossible to trace one firm's influence on the final outcome.
Think of this as chain drain: each link in the supply chain creates an opportunity for optimizing the cost but comes at the expense of learning.
Simple buildings can survive chain drain and be more or less OK. No one’s worrying about the big box that Target’s housed within. But complex or high performance buildings either suffer in quality due to chain drain or become crazy expensive as the assembled team work to apply collaboration and project management like so many fingers in a leaky dam.
To put some scale on the cost, Apple's Norman Foster designed campus came to almost $1,800 per square foot. This is 2-3 times more than typical class A office space. The average US home built at the same cost per square foot would be $3-4 million.
There are no public details of how this money was spent. Sure, some of the cost is due to indulgent things like gigantic custom glass panes, but it's almost certain that the coordination of the project was also a significant “tax” paid in the form of project manager salaries, mockups and prototypes being built to align all of the trades, and other investments in clarifying shared purpose and desired outcomes.
Chain drain should matter to people who care about data because every drop of chain drain is a bit of data that goes uncollected and un-utilized.
Chain drain makes buildings more expensive and less performant.
The horizontal complexity of the built environment's supply chain makes it possible to built complex stuff at great expense or build simple stuff at moderate cost.
Among efforts to innovate within the AEC sector, there's a preference for "prestige" technologies like BIM and VR that promise to radically change the way a firm does its job. You can see these fascinations in the chart below. What's missing from this chart is the boring workhorse software that has driven huge productivity gains in so many other sectors: business process automation and enterprise software more broadly. Here's what it looks like when visualized in a paper by Brozovsky, Labonnote, and Vigren:
What chain drain tells me is that the barrier to higher quality buildings and spaces is not making slow firms work faster, but making a leaky supply chain work better.
The most extreme version of fixing the supply chain is a company like Gropyus. You point them at a plot of land and tell them how many units of housing you want, write a very large check, and they give you a building. Designed, built, and managed via custom software and hardware.
In their model, the distinction between architect, engineer, construction manager, and contractors are all subsumed into a vertically integrated business. They sell a bulding as product.
That's how 45% of housing is built in Sweden, with similar numbers across northern Europe. Japan comes in at 15%. Making such a dramatic change requires, in the estimation of Ivan Rupnik who is a leading scholar of this subject, a customer to guarantee purchases at scale large enough to justify building the factory. That's often a government, and thus unlikely to happen in the US any time soon.
Less wholistic examples of supply chain reform are exciting too.
Founded in 2019 by Federico Negro, a Wework alum, Canoa helps people do interior design. Instead of speeding up a clunky horizontal process that is spread between Autocad/Revit for drawing layouts, PDFs for product information, spreadsheets for keeping track of orders, and presentation decks for getting buy-in. Canoa provides a bespoke tool to do all of this.
In other words: it manages data across a handful of steps in the process. And in the process makes it easy to give the user continual feedback: here's how much this design is going to cost, here's how much CO2 it will cause.
It's like a portal gun that let's you jump from sketching a design right to moving in, skipping the annoying stuff that’s usually between A and B.
Canoa's core product is a beautiful, fast, and multi-user (like Figma). It's also opinionated, meaning the tool makes certain kinds of use easier. They've built Canoa to encourage you to make choices that foster a circular economy.
Canoa's product follows the logic of the internet and does it's part insanely well but/and is interoperable with other tools by encouraging data to be imported (e.g. grabbing furniture info from the open web) and exported (e.g. to industry standard tools).
It's a product-based business, obviously. Canoa makes their product—the platform—and can spend all day, every day worrying about how to make it better for their users. They do not also sell design services, which would both muddy their business model and fracture their attention.
In an alternative reality Canoa is built inside a global AEC firm as a secret weapon that allows their employees to design interior build-outs faster. This is less plausible because: 1) AEC firms generally lack the software development expertise needed to deliver at the level of Canoa and 2) their financial structure and lower margins tend to make them hesitant of the up front investment that is needed to build such a product.
Who will win? Will AEC firms figure out how to rewire their sector and produce better outcomes, or will outside disruptors use software and data expertise to delivers the same outcomes through entirely different means?
I’m feeling pessimistic about firms.
Products—and product based business models—are a form of leverage that lets experts have an outsized impact. We hope the impact positive, of course.
The neat thing about building opinionated products is that it's a way of shaping future outcomes, such as more workspaces designed from a circular perspective, without requiring massive training of everyone involved.
If you really want to put data to work, own the business model, think in products, find a niche, and care more about the supply chain than disciplinary boundaries.
For prior musings on this topic: see Scrambling up Relationships in AECO and What Happens when Buildings Become a Product?
These weeks: Job talks. Cities intensive. Studio reviews. Thinking about Fall already. 🏃