#3 - What Vertical Integration Means for Architects

Also: College in a post-pandemic world, worth it?

Hello and welcome to new subscribers! Two big topics on my mind for this issue. If you’ve stumbled on this and enjoy what you read, please subscribe so you don’t miss future issues!

This week, two relatively new entrants into the pre-fab residential startup space received some good press. The first is the launch of Juno, the latest Silicon Valley start-up to try to tackle pre-fab housing. The second is a great article about a company called Apt: The First Natively Integrated Developer, written by Packy McCormick for his Not Boring newsletter. Both companies are riding the wave of an ever-growing list of companies that are trying to integrate and consolidate different elements of the building industry into new business models.

There’s a lot of really great stuff in McCormick’s piece, but a couple of specific passages really clarified for me some concepts that I have been mulling over for a long time.

The first is this:

A building has two core customers: its end users (people and organizations that occupy the space) and its financial backers (real estate investors and lenders).

With single family homes, the end user and the investor can often be the same person. [But…] for multifamily residential buildings, there’s almost no overlap between end users and owners. 

Not only is this true in the residential market, it’s even more pronounced and fragmented in the commercial sector, where it is common to build a “core-and-shell”—essentially, a giant empty building—that then gets rented out to lots of individual tenant companies, each of which may go through their own fit-out of the space, kicking off a separate sub-project for each occupier with its own project lifecycle, and each of whom likely have an in-house real estate team that is managing the build-out on behalf of the final set of employees who will actually inhabit the building day-to-day. In this way, the ultimate end-user of a building can be many layers removed from the original designer of the building.

The second interesting point highlights the potential benefits of stronger vertical integration across the building industry. As McCormick writes:

Over time, by repeatedly treating every development project separately, the industry as a whole moved towards extreme degrees of vertical (design, construction, operations) and horizontal (specialization by trade) fragmentation. Every part of the building value chain from finance and legal, to architecture, engineering, construction, and property management has been split.

Again - this generally makes sense for each party involved! But disintegration has created a barrier for innovation in the industry as a whole. To coordinate design and construction, all industry players rely on often outdated standards and processes (some decades and even centuries old).

There’s been lots of analysis in the tech industry about the benefits and challenges of integration vs. modularization; Clayton Christensen based his whole Theory of Disruption around this topic. From the article again:

Looking at the real estate development value chain, Apt is the only company that integrates investment management, development coordination, and design and engineering, while modularizing (no pun intended) manufacturing and construction.

From Apt: The Natively Integrated Developer by Packy McCormick

I love that diagram, and its a great framework to start evaluating these different companies, but I think it’s missing one additional final column that would have really tied the two themes together: Building Operations/Facilities Management, which represents all of the products and services that are required post-construction, when people actually move in to the building. In theory, by taking on the challenge (and risk) of actually running the building, there’s an opportunity to deliver a better, more unified user experience, and also capture more of the value (think of how Apple owns both hardware and software on an iPhone to provide a tightly integrated ecosystem, and how they are able to charge a premium for it). There are a number of companies that directly serve the end user in exactly this way by managing the operations of the building post-occupancy, and it would have been valuable to include them on the chart.

To McCormick’s (and Apt’s) credit, he does hint at this later in the piece:

To start, Apt is focusing on investors, but the relationship with the renters who will live in each building is no less important for a [Natively Integrated Developer]…For a NID, it’s crucial to learn from the end-users of a building and create a feedback loop that continuously improves their product.

From this perspective, I would classify a company like WeWork as having integrated the Development Coordination, Design and Engineering, and Building Operations stages by not only finding and building locations but also providing value to its community members after they moved in, while modularizing the Manufacturing and Construction stages by partnering with external vendors and local GCs. By operating the building, we were able to create a strong sense of community and learn from our members, to inform future designs and provide the feedback loop of learning to improve our offerings.

(Towards the end of my time at the company, we were in the midst of integrating Investment Management with its Ark fund, Construction through the acquisition of contracting companies and self-performing construction on more locations, and even a little bit of the Manufacturing as well by fabricating aluminum storefront and some of the furniture on our own. Perhaps there is a breaking point where one can take on too much, too quickly. You need flawless execution to pull off so many different activities well, and we were certainly far from flawless operationally at WeWork.)

The Un-bundling and Re-bundling of Architectural Services

I want to zoom in one particular part of the building process that I know most well, and than is architectural design. In the diagram above, traditional architecture firms live solely in the stage of Design and Engineering, and is stubbornly stuck in the lane of Bespoke Buildings. Practically, what this means is that, of the two customers of a building (financial backers vs end users), architecture firms have traditionally been farther removed from the end users, who exist in the (imagined) 6th column of Building Operations in the chart, and much closer to the financial backers, who live in the immediately adjacent Development Coordination and Investment Management columns and who presumably are often the people who actually hire the architectural services.

The introduction of more companies integrating the design stage directly present a unique opportunity for architects to serve end users of their buildings more directly, particularly if the company is one which has opted to operate and manage the facilities once its actually inhabited. Traditionally, this has been limited to certain large players in the hospitality (i.e. hotels) and retail industries, but more and more companies, from real estate developers to co-working companies, are employing architects in-house rather than outsourcing that to traditional firms.

There’s likely good economic and historical reason for why it took so long for architects to make this move, but integrating architectural services into a combined bundle of offerings can have profound effects of the actual practice of design. For one, it aligns the interests of the architects with the interests of the end user of the space, something that was shockingly lacking in traditional practice.

Ask any architect in a typical firm about how a particular mixed used complex that they may have completed six months ago is doing today. How do the inhabitants enjoy the space? Is it occupied and fully used daily? What aspects of the design work well, and what could have been improved? Chances are, they won’t be able to answer you, because there’s little financial incentive to do so: no one will pay to have an architect to return to a project to analyze performance, and even if they did, the chances to reapply those learnings on a future project is limited when each building is bespoke.

I’m a product manager today, and in the tech world it is so valuable to be able to add instrumentation to a software product to understand how a piece of tech is performing, to analyze a design and A/B test different iterations, to measure how it impacts KPIs, and to let data drive decision making. None of these techniques are common in architectural design; in fact, the prevailing attitude may be the opposite—architects are gifted with preternatural instincts and should be able to use their design sense and intuition to make decisions. What most excites me about architects being embedded within a company is to reconsider to whom they are delivering value: ultimately it should be about the end user experience, not the financiers. (If you keep the end users happy, money will flow back to the developers in the form of positive cash flow and increase in building equity, trust me.)

There’s another parallel to be drawn here between companies that do business with other businesses (B2B) vs. consumers (B2C). The defining element of a B2B product is that, unlike in a B2C product, the buyer and the end-user are not the same.

In many ways, traditional firms have been working with (and for) developers and financiers, and far less so the end-users. With the introduction of these new integrated business models, we are starting to see that the role of the architectural designer is now being reoriented from a B2B model—catering to those who hold the money and are making the buying decision—to a B2C model, where the focus is on providing a great experience to those who actually inhabit the buildings.

The Software Opportunity

This shift also has an impact on the tools and technology that architects use to design. Existing software has been developed and marketed for the old paradigm of vertical and horizontal fragmentation. There are tools and platforms that target each individual functional layer of the stack: real estate software for managing portfolios and transactions, regardless of the sector; 2-D CAD drawing and 3-D BIM modelling software for architects that can model just about any building imaginable, provided you try hard enough; construction management software that provides a one-size-fits-all solution for all sizes and types of contractors. Each of these try to appeal to the broadest swath of users in a particular stage; few of them integrate well with other software that exists across the boundary of a stage.

Take the current industry standard for building modelling and construction documentation, Autodesk Revit. In trying to be a comprehensive building design suite all parts of the design process, it has become a bloated piece of software that does no particular thing well. There is, for example, a building massing tool that is great for early experimenting with building envelopes and form; this is useful for the small segment of the market that does ground-up, large scale development on multi-story, multi-unit buildings. But for an interior designer that focuses on home renovations, this is a feature that will likely never be used. Meanwhile, Revit is severely lacking in tools and good user experience for managing things that do matter for interiors, such as working with materials and finishes or modeling small objects with the right level of detail. Revit completely overserves a large segment of the market along one dimension, but completely underserves it in another.

Each company I’ve talked to that has moved into this integrated model has also had to severely adapt existing tools to their unique needs, as well as develop custom solutions to sharing information across different stages of the process. Transferring relevant data from real estate to design to construction is no easy task, because the lack of integration means little standardization of data formats and application interfaces. (This is why the landscape of design technology is littered with interoperability tools that try to facilitate moving data between platforms.) Whether it’s developing plugins for existing platforms to import and export data or creating customized project management software and workflows, this can be an expensive and time consuming endeavor—the type of investment only makes sense if you can amortize the capital cost of software development over time by not treating each project as a bespoke one-off.

As more of companies develop around vertically integrated models, the opportunity opens up for more niche offerings that will target a specific vertical market or typology. An end to end platform for developing, designing, and constructing single family homes may end up looking very different from one that is designed for civic works that target local governments, or commercial office buildings.

(I’ve built a career around customizing and building software for these types of integrated design companies. I have lots more to write about on this topic in upcoming issues.)

Bootstrapping the College Experience

Much ink has been spilled already about how the future of the office is up in the air in a post-pandemic world, but we’re also starting to see how other institutions such as colleges and universities are re-imagining the educational experience come the fall.

Though the majority of schools still intend to host in person classes, a number of schools are also experimenting with different hybrid plans. This week multiple universities announced their plans about how they plan to continue offering courses (often virtually), while also limiting the number of students on campus at any given time, often by staggering which classes will be allowed to return to campus. My own alma mater, Yale University, is alternating first-year and sophomores’ returns to campus while allowing juniors and seniors to stay the whole year, while conducting most lectures and seminars remotely. Here’s a brief roundup from the New Yorker of the variety of approaches:

A majority are planning to hold classes in person. Some, like Middlebury, a small liberal-arts college in rural Vermont, and Bard College, in upstate New York (where I will start teaching in August), are planning for most students to return to campus, and provisionally committing only to minor adjustments to the academic calendar. Some, like Bowdoin College, in Maine, and Amherst College, in Massachusetts (where I have been teaching), are reducing the number of students on campus and providing everyone with a single dorm room, and will transition to remote learning after Thanksgiving break. Others, like Oberlin College, in Ohio, are radically changing their academic calendars, switching to a three-semester, year-round schedule to insure that no more than two-thirds of students are on campus at any given time. Stanford will bring half of its undergraduates back to campus at a time, but it plans for most classes to be taught remotely, as does the University of California system. Harvard is bringing up to forty per cent of undergraduates back to campus, where they will take classes online. Some classes will meet outdoors—Amherst, for example, will erect tents for outdoor instruction, which will allow for social distancing. All schools are putting testing and contact-tracing protocols in place: Yale is planning to test students weekly, Bard will require students to arrive carrying negative test results, and Amherst is setting aside residence halls where students who have tested positive will be quarantined.

One might expect that the cost of tuition may be appropriately altered in light of such drastic changes, but Harvard raised some eyebrows this week when it simultaneously announced that all of its classes would be moved online, but that its tuition would remain at $49,653, the same as it would be for an on-campus learning. The university, which has the largest endowment of any university in the world, was appropriately dragged on social media for the decision and left many concluding that not even a Harvard degree was worth that amount of money for eight months of sitting in front of a screen that could be done anywhere.

While I think that it’s premature to pronounce the “death of higher education” (though online education startups and coding bootcamps have been championing that idea for years now), I do think that the new policies in place for the fall so severely cripple the undergraduate experience that it calls into question whether the cost of going to this upcoming version of college is still merited. So much of the value that I got out college was due to in-person interactions—even more so than for work. From extracurriculars to dining hall conversations to dorm room parties, the value of face-to-face experiences were critical to my personal and academic growth, and it’s hard to imagine my college years without them.

That being said, I have no doubt that the youth of today can come up with all sorts of creative ways to try to replicate this experience. Gen Z is already adept at leveraging social media and online platforms to facilitate connections, express themselves creatively, and engage with each other from a distance. The question, then, is whether or not paying for a college experience that is stripped of face-to-face interactions is still worth it when there are so many ways of replicating the benefits of that, for free, using existing platforms.

If I were an incoming freshman, I would be extremely skeptical of the value proposition of what is being offered by universities for this upcoming fall. I can imagine a world where the incoming classes of each school can come together to recreate an experience that very closely mirrors (if not surpasses) what I anticipate college life to be like in the coming months. Here’s how:

  • Leverage existing social networks to reach fellow accepted students. There are lots of ways to build a community online, but if you’re looking for a level of exclusivity, well, the admissions committee has already done the hard work of vetting each student though the application process, you’ve just got to seek them out. Thankfully, accepted students tend to find ways to congregate anyway: the Yale Class of 2024 Facebook group already has over 600 people, and an established process for vetting that each person is in fact an accepted student.

  • Take online courses for free. Top-tier universities have long offered free, online versions of their courses, especially for the larger intro survey ones. The content of a liberals arts education probably doesn’t change very much year over year, so the materials are pretty evergreen, and the experience is likely to be very similar to exactly what the schools plan to offer in September anyway. Platforms like Coursera or EdX even offer different forms of assessments, to validate and prove that you’ve learned the material.

  • Set up long-lived channels for communication. A Slack space or Discord server with dedicated channels or rooms for each course or topic of interest. You could easily have online discussions and collaborative homework help on any of these platforms, break out easily into video chats if necessary, and generally

  • Discover what extracurriculars operate well in a digital world. It’s hard to imagine things like athletics, theatre, and performing groups adapting well in this new environment, but many other activities can be set up and I’m fairly certain that universities are banking on the ingenuity of the students to come up with creative ways to solve this, rather than offering any solutions of their own. Student-led organizations were always lacking in administrative support and funding, why expect that schools would start now?

To be clear: even suggesting this as a viable alternative presumes a certain amount of privilege that students’ home lives can support this model of learning, or that anyone can just afford to take a gap year or get a paying job overnight when 40 million American adults are out of work. Universities play other roles than just offering classes; they also provide safe spaces for free thought and facilitate upward mobility for students and their families. This morning, the NYTimes went in-depth into the challenges that first-generation, low-income (or F.G.L.I.—new acronym alert!) students, who often rely on the physical and social environment of an on-campus college experience, are facing trying to compete for the limited amount of dormitory rooms that Harvard is offering. As a F.G.L.I. college student myself, I’m not sure that I could have pulled off what I’m suggesting above fifteen years ago as an 18-year-old eager to escape my inner-city Brooklyn neighborhood.

Nevertheless, college was always the time when you transitioned from an educational model that kept you in a classroom for 6-8 hours a day during high school into one where you maybe had 10-12 hours of total actual instructional time per week, and then had to figure out how to best fill the rest of your time. Part of what accelerated that personal transformation into an independent adult for me was social proximity to others: making sure I did the readings so I didn’t look like an ass in a discussion group, working on a problem set in person together, presenting my architecture studio work in front of my peers, figuring out when and how to clean my room and do my laundry so my roommates wouldn’t lose their minds. By the looks of it, college in the fall will offer little to none of these social signals that drives personal growth; instead, students will need to figure out ways to go through this largely on their own. I’m optimistic that they can figure it out, and I’m excited to see how they pull it off. But why pay $50k for the privilege of the experience, when you’ll be expected to do so much of the legwork yourself, anyway?