The pace of change in the AEC industry has accelerated. Automation is surging, and a growing number of start-ups with innovative business models are entering the market. There’s a whole new vocabulary of tech-enabled service providers: PropTech, ConTech and BuiltTech. What is driving these disruptions and what future should we bet on?
For the 2017/2018 Thornton Tomasetti annual report, we convened a group of specialists to discuss the challenges of disruption in AEC. Here is an unabridged version of their discussion.
DENNIS The impact of technology on business and culture in the AEC industry is clearly accelerating. “Disruption” has a flavor of Silicon Valley in terms of technology’s impact on business models and of capitalizing on and monetizing radical change. Can we sort out where the AEC industry along that spectrum of change?
CORY I’m interested in data governance and how we could capitalize on use cases beyond providing the services of design and construction. If you look at outside disruption and the creation of digitized value chains, the architecture itself — the design of a space — would have nothing to do with it. It’s what we can gather from the use of buildings that becomes beneficial to our industry. Companies like Amazon and Google are already tracking what we do and say through Echo and Google Homes. How can we benefit clients and monetize those services post-construction? This is where outside industries see more value in AEC than we do, as architects or designers. The future is less about brick and mortar and more about people and data.
RICCARDO We’ve seen quite a bit of evolution in this industry over the past ten years, but it hasn’t been a revolution. It’s been about making better tools, becoming more efficient — integrating technology to automate the objective parts of our work, like using parametric modeling to come up with a better core, or typical column sizes, or the best sustainability strategy. We’re all about optimization, but that doesn’t help us answer subjective questions: “Is this a beautiful building? Will this make a great place? How will it impact the community?” As designers and engineers, we’ve seen our work commoditized and its scope reduced. Now there are contractors and developers who can do what we do. Here’s the scary part: The New York Times recently ran a story on an AI facial recognition program. By learning certain patterns in a few days, the program was able to construct a beautiful face, based on millions of data points. Researcher put these constructs next to photos of real people and asked respondents: “Can you tell the difference?” And they couldn’t. This says to me that at some point you’re going to be able to push a button and not only optimize a core and the building envelope and the best sunlight strategy, but develop a beautiful building. I don’t know about you, but I’m an architect and I’m asking myself: “Why do they need me?” So the disruption I worry about isn’t monetizing what we have; it’s something new. Take Lawyers. They traditionally developed an understanding of law through reading contracts – a long, slow grind, just like we did, working on construction documents. Lawyers don’t have to do that anymore because AI reads those contracts, a thousand a minute, and summarizes them. All of a sudden the practice of law is radically different. That’s coming our way to change the subjective part of our design.
BRETT No one has shown the path of disruption in AEC but once it happens, there will be a rush of fast followers.
KP In our industry we tend to forget who the customer is. Over the past five years I’ve talked to more than 100 owners and they don’t care about Revit or parametrics. They don’t care about you doing your job better or more efficiently. That’s your problem. When designers or the built environment community come up with something disruptive, they are not able to demonstrate its value to that owner. “You have BIM? Great! Now I can pay you 10% less. Robotics? Terrific — I can pay 50% less.” Give me real data that tells me that I can build a building cheaper, better, one that performs over the long term. Most designers and contractors can’t even tell an owner what their operational costs will be for the next five years, forget about 15 years. We get caught in our own brains around innovation, disruption and what it means to us. But what matters is what it means to the owner.
ZACH Let me push back on that. As a client of these types of firms, we send a lot of technology platforms to the people who are doing the work for us. We’re not withholding interesting articles about innovation. Our attitude isn’t, “Let’s replace these people right now.” It’s: “Let’s help them, and maybe we’ll save money and get higher efficiency because of it.” Fundamentally clients want their buildings built more efficiently, so they want to enable the consultants as best they can, whether that’s the GC, a geotech, or the architect. I introduced our architects to Envelope, which was incubated by SHoP Architects. Handel Architects and SHoP embraced Architizer early on. That could theoretically disrupt the big thing that you need architects to do, like write specs for materials. Innovation is a push and pull. There are definitely clients who love that you all are embracing technology, and it’s not necessarily to destroy you in the long term. It’s a short-term objective of building more efficiently.
KP Agreed. I do think, however, that most owners are frustrated. It’s not: “You’re doing an amazing job. Let’s do better.” It’s: “Come on, guys, you just faxed a doc.” I’ve had a few owners say, “You know, sometimes we feel we’re condescending to our design-build teams, because we’re like, ‘Hey, there’s this thing called the internet. Maybe you should use it.’”
DENNIS What opportunities for new services or business models are emerging in this changing landscape?
CLAUDIA I represent a fabricator working with architects, engineers and developers, with whom we have great relationships. We’re the last link in the chain, working directly from models, and do everything in-house: design, detailing, fabrication, and coordination. We believe in transparency with all team players to build trust. No one else taking this end-to-end approach. In what we do, experience makes the difference. Many people who have been with our firm for 30 years know, for example, how to heat steel to bend it. That’s an added value that comes from collective experience. So that makes us really different from anybody else. Loyalty counts since it concentrates knowledge and makes it available. Ten years ago, we didn’t have relationships with developers or architects. We had relationships with contractors or subcontractors. And then an internal integration of services — consulting, design and fabrication – required new relationships.
DENNIS Yours is an example where new opportunities emerge out of evolving roles and relationships. Developers must be seeing a value proposition in your approach. What do you do next and how do you stay ahead of the curve?
RICCARDO In our business, technology has replaced a lot of those guys who drafted by hand — our version of Claudia’s people who know how to bend steel. Software pretty much outstrips their ability to do that stuff. You can entertain more options faster, but that doesn’t always lead to better buildings. Documents on buildings from the 1950s were 50 pages. Now they’re 5,000 pages, and a bunch of 3D models. Just because it’s possible doesn’t mean we should do it – especially if it leads to some terrible buildings.
BRETT Whenever you have disruption, the consumer wins. Think of Airbnb, Uber, digital cameras. The example we use is: How many hours does your firm spend designing bathrooms?
RICCARDO I don’t know — a lot.
BRETT How diverse are bathrooms? Always the same thing. You can write a configurator for bathrooms. And then brand the bathroom, right? A Banana Republic bathroom. It’ll be fantastic, right? And ten 10 people worldwide will manage the design-production of Banana Republic bathrooms, so we can take that money and put it into more interesting steel structures. Or return it to employees or shareholders. If you build a bathroom configurator, and a classroom configurator, a staircase configurator, and an electrical-room configurator, pretty soon you have a stack of technology. Then you go to a developer, and say: “I cost 10% less, and the quality is still there.” Now you have a competitive advantage, and you go attack a vertical. That’s Pixar.
KP But I don’t want a repetitive bathroom.
BRETT Technology can add the diversity and mass customization.
CORY We’re trained to think of a bathroom as a standardized commodity, but that isn’t necessarily the case. To build a technology stack that standardizes what we believe to be best practices is just automating an old set of processes and ideas. That’s not using the technology to improve our work. For example, when Google’s DeepMind first programmed AlphaGo, they trained the AI based on historic games. During subsequent games against professional Go players it made moves that were assumed to be wrong, but the AI ended up winning anyway. Why? Because the priorities that the computer created were different than what we always thought were the right priorities of the game. And when they created the second iteration AlphaGo Zero, they let the AI train itself. It ended up winning against its original version, 100 to 0. The point is, we have ingrained priorities that we assume are right but a computer is not necessarily going to think that way. So we have to look at how we build data structures and classify information. A bathroom may not have to look like what we think a it “should” look like. To fully leverage new technologies, like AI or ML, we must get rid of history to some degree, because it’s built on assumptions and processes that we evolved based on code or pro forma or what we thought was the right way to deliver a project at a given time.
KP Do you notice a shift in attitudes about sharing? I remember my first CRM system and thinking: “I’m not putting my customers in there; they’re mine!” There’s a generation coming up now that believes that if I share everything, everyone benefits.
DENNIS Is there an evolution about trading control for speed, particularly around intellectual property?
CORY In other industries, patents are like baseball cards when we were kids: I’ll give you a Rod Carew for a Keith Hernandez. They’re trying to get a complete deck. Companies aren’t patenting to lock things up but rather to create a more collaborative environment. In AEC, it’s another story. A small number of companies own large portions of the market and they don’t necessarily want to share because there is no business case for them. Within most architecture and engineering firms, and especially with our vendors, data is still thought of as proprietary and seen as a business advantage.
CLAUDIA We don’t follow that practice. We share everything because when we do, I can say, “you know everything I know” and that leads to a better outcome. I’m not afraid to share anything with anybody because, at the end of the day, data is only data. In my field, experience makes a big difference. It’s what you do with data that matters.
RICCARDO We look at patents differently. We want to lock up the IP for a little while and see if licensing agreements will make money while we’re sleeping. But it’s hard because we’re all professional service people, and we’re asking: “What are we doing in the IP business?”
DENNIS Products are hard because they take investment. They take time. They’re big bets. It’s not the traditional DNA of service companies, although I think that’s changing. I also think cost and other barriers to entry into developing software are dramatically lowering.
ZACH The most attractive companies to venture capitalists in the real-estate tech space in the past 18 months have been all tech-enabled service companies: WeWork, Compass, Redfin, Katerra, Opendoor. I think there’s a chance the industry will spawn even more. WeWork is trying to get into construction vertical integration activities. Convene has a fully integrated design-build team. When you have homegrown IP, you have to decide: Should I kill this thing; keep it in-house in core services; spin it out as a SasS [software as a service] product, or as a tech-enabled services company?
BRETT Or think about how that IP can enable you to do a little bit more and create the opportunity for more valuable innovation.
ZACH In our accelerator program last year, we had a company called Bowery doing commercial appraisals—a niche but large market. We went on a road show to potential customers, mainly banks and said, “Buy this software,” and the banks said, “This is amazing,” and never bought it. So we said, “OK, let’s pivot and become a tech-enabled appraisal firm ourselves.” I think you’re going to see more of that — companies that are trying to sell into your industry get frustrated, and ultimately form their own business.
KP The move from services to products is a big shift. In professional services, sales and marketing supports you. They help you with your RFPs. They’re not out there generating leads, talking to the market. It’s the architects and engineers who do that, and they return from a meeting and say, “You know, we should launch a new service. Hey, get someone in Marketing to do a brochure.” The product business is the opposite: marketing drives the bus. This is a challenging for architects and engineers who are used to driving the bus.
BRETT Innovation within our space involves big risk, because any time you change how you normally do something, you’re asking your client to trust you.
ZACH I’m encountering this now. We’re shopping around a company that does product progress documentation, and we have to make a four-headed hydra sale: We have to sell to the developer, the GC, the CM, and the insurance company. They all have to push on each other because they all want it, but they’re all scared of it for different reasons. There’s a lack of understanding of who’s going to have access to what and how it’s all going to work. This is a big hurdle for companies looking to sell products into the industry.
RICCARDO As Cory mentioned earlier, our most valuable product may not be our designs or IP but our data. Maybe the suitor who wants to buy your company or mine is an aggregator of data. Could be Google or IBM. “Let’s buy up all your data–not because we want you as engineers or care about you as architects. We just want your data.” They’ll figure out its meaning and use it as a platform for themselves.
BRETT We have a client who is working with a big city to locate a bridge across a river, and they’re doing it using social media geolocation to track where people are. They do this for stadiums too. So part of the IP at an architecture firm is all the products that are specified, and the interior design. I can probably mine that from Instagram photos of corporate environments or from your website.
DENNIS We’ve talked mainly about outside forces acting on our firms or our business models. Let’s look inside our firms: How do we become self-disruptors? Do you see that as an imperative?
RICCARDO My partners and I talk about this all the time. We know we have to shake things up. A lot of our time is spent on how to make our business better — while handling 200 emails a day. How do you keep going down the freeway at 100 mph and change the tires and paint the car? I don’t have an answer. But I have a technique to get there. Business schools these days are big into design thinking, which is what architects have been doing all along. The architect starts with a clean sheet of paper and tries to solve a problem that hasn’t been solved before. So I’m very optimistic about our ability. That’s the good news.
CORY I see two approaches to technology in our industry. The first is addressing inefficiencies in design and construction processes. In most cases this results in incremental changes that provide slight improvement. But we’re typically interested only in the direct outcomes: “The project is done, let’s move on.”
The second approach has a long-term goal of improving the entire project lifecycle. Through the application of data-driven processes, firms are looking to improve project outcomes by continuously assessing designs from early simulations through post-occupancy performance. This process allows us as an industry to create digital value chains that inform new development through the real-time application of building information. This approach is moving in the right direction. To self-disrupt, however, architects and engineers need to start thinking about how we monitor, maintain, and support a global infrastructure rather than just how we design it.
KP When people ask me, “What can we do to self-disrupt?” I tell them step one is compensation. Right now the people at the top are motivated to sell their stock to the people in the middle. That’s the succession plan. The disruption we need is to rethink this traditional compensation model. This means paying people at the bottom based on their talent and their market value, and that’s not $75,000 a year. The new CTO of Walmart makes 10 times more than the CEO. The CEO is a commodity, the CTO is not. In AEC, the system is based on seniority and pay bands and whether you get your license, which is a broken system. If you really want to disrupt your organization, you say: “The people at the bottom who are doing innovative, disruptive things need to make more than the CEO. Period.” That’s where I would start.
BRETT In our business the most exciting thing is tracking how people build buildings now. We have a clash-detection tool that allows us to track every element that’s been modified, added or deleted by every user on a project, with their name, the time stamp, what room it happened in and what they did. That big data lake tells us how a building will go together. When we add automation, we’ll have a bright red line showing how we’ve deviated from baseline and how we’ve improved. It can be pretty disruptive to tell someone, “Hey, you’ve moved a door five times.”
DENNIS What are the disruptive impacts of version tracking on the firms that you guys are working with?
BRETT The people who don’t want to be tracked will leave. The people who perform well will get paid more. We can define and characterize owner change in a data-driven way. When you add innovation, you can track the value of the innovation. We tend to think of innovation, like automation, as being about a robot or a computer doing a thing, but that isn’t how it works. A person will use automation, and they will get credit for the automation as it’s applied to a project. That person gets the value of the automation added to their bucket of how they’re working smarter and better. That’s how we see it in the near term.
In my world — and perhaps for all of us — there are two things that are more important than patents: customers and data. As a firm you get to keep your data. That becomes what you do. And you can use that data to your advantage. That may be the next big disruption.
Follow these topics: