Episode 23

Network Science Applied to Workspace

Two workplace architects/product/curricula/software designers, interested in the application of network science to information access, urban policy and planning, and the built environment, talk about their research, think through some complex theories, and contribute knowledge to improve workspace design output. Their conversation will create an appetite to learn more.

Subscribe on:

Featured on the Show
I think we just wanted to open a dialogue with a broader community and introduce some concepts with which many design practitioners might not be as familiar. I mean, network sciences, we all live and participate in networks every single day. But understanding the behavior of networks is potentially esoteric to the design community. Josh Emig, Product Leader | Architect

Transcript

CCB: [00:00:00] Welcome to the ONEder podcast, this is CCB, your host, and I'm here with another one of our 2020 ONEder Grant Award teams, I'd like to give you just that bit of information about the ONEder Grant, which is the ONEder Grant supports exploration and research that impacts environments where work, innovation, healing, and learning take place. And we have given out nine ONEder Grants over the past two years. Today we're talking with Nash Hurley and Josh Emig and I'd like them to introduce themselves and give a brief overview of their topic title.

Nash Hurley: [00:00:53] That's me. OK, so, I'm Nash Hurley, I'm an architect. I've known Josh for the better part of two decades. He and I met in New York working for kind of a standard core and shell architecture firm. But since we've gone on and found ourselves working in workspace design in different capacities, Josh, a little bit more on the product end, and me on the strategy and product side of things. The title of our grant "Network Science Applied to Workspace," over the course of the grant really probably became more workspace as a link and thinking about workspaces, as a link and then the network of production. So I don't know, CCB, if you want me to go into why we chose that particular topic or if you want me to unpack that a little bit further.

CCB: [00:01:53] I'm going to have Josh introduce himself and then I'm going to ask you to unpack that a little bit further.

Josh Emig: [00:02:01] Hey, I'm Josh Emig. As Nash mentioned, we had been friends for quite a long time and former colleagues, and we started talking last year before the pandemic about some ideas that we wanted to explore together. At the time, I was working as a product director, working on software at WeWork, where I was the Head of Research and Development prior to the product development work that I did. And we just had a similar line of thinking just in terms of where we saw the opportunities for work, and the future of work and cities, and the ideas that we thought would be, you know, that would shed light and help us to flesh out some clear, let's say, clarify our thinking on those topics by using the network--the lens of networks.

CCB: [00:03:00] So you've explained why you chose the research topic, but can you explain, what did you hope to accomplish through this study?

Josh Emig: [00:03:13] I mean, for my part, I think that my hopes were pretty simple. One, like I said, I wanted to clarify my own thinking on the future of work in cities. These are topics that were very near and dear to a lot of the work that I did at WeWork. And even prior to that, working with some of the workplace designers and some of the technologies that I that we developed at Perkins and Will. And so it was just kind of an opportunity to clarify that, obviously a lot of things have changed in the world. And last year was kind of a doozy. So there was that. And then also, I think we just wanted to open a dialogue with a broader community and introduce some concepts with which many design practitioners might not be as familiar. I mean, network sciences, we all live and participate in networks every single day. But understanding the behavior of networks is potentially esoteric to the design community. And then lastly, I think for me, I just wanted to take advantage of the opportunity to gain some depth in an area of knowledge and in which I had an interest but a rather cursory understanding.

Nash Hurley: [00:04:36] Yeah. All that makes sense to me, I think also for me, I was looking for like a really for like a tool and some better language for engaging with large project teams. I work on pretty large, complex projects. And I realized that we didn't have a lot of common language around some of the business needs that our clients had as it related to workspace. We would start with a solution. We talk about adaptability, we talk about flexibility. But we didn't have a framework to kind of hang investment in those things on. So it got kind of a value, a lot of decisions get evaluated against potentially like older models of business, and I was looking to have a framework that could potentially unify some of the language between economics and workspace, to have more business focused decisions around investment and workspace.

CCB: [00:05:34] So I am going to point out for anyone listening that an abstract of this particular project topic is available on the One Workplace website as well as the full research paper. And when you, if you have time to look at it, it is organized extremely effectively. And I'm going to give you kiddos for that, gentlemen, because it is a simplistic approach to an incredibly complex topic, the way that you have broken it down. So I kind of wanted to ask that question about, for you working together remotely, what form did your research take?

Josh Emig: [00:06:20] Well, I mean, we really executed the research almost pretty much in parallel to the way that the resulting deliverables were structured in which, we did a literature review like you do, and we read a lot of books and did a lot of studying and shared a lot of papers and did a lot of discussion about that and sort of formulating some early ideas about how this theory might be applied to the workplace. We reached out. The second thing we did was case studies with, I'm sorry, not case studies, but interviewing some subject matter experts. So we reached out to a few people that worked in hospitality, a few others that were sort of prominent figures in organizational network analysis like Ben Waber at Humanyze. We reached out to César Hidalgo who wrote a great book a few years back called "Why Information Grows" and uses both a framework of sort of a combined framework of physics, economics, and network science to describe the way that economies work, which is a fascinating book. And so we had a chance to interview him. And then we went on to do a much more in-depth case study with Wikimedia Foundation. And so, which really sort of took a lot of the things that we had synthesized through the literature review and learned from our subject matters and sought to sort of use them as a lens to understand the way that that particular distributed organization was functioning.

CCB: [00:08:01] Did you want to say something there?

Nash Hurley: [00:08:04] I always want to say something, but I think that Hidalgo’s book about "Why Information Grows" is really foundational for advancing my thinking, really. Some of the principles in there around transaction cost theory and what that means for spaces of work. A lot of his focus was more like manufacturing spaces. But what was really interesting about the network science is that it could apply to both. I think the other thing, just to point out, that came to mind as Josh was talking, is how our research actually flowed upon our own connections. So we reached out to subject matter experts that we had known from past projects. We happened to have done a project for the Wikimedia Foundation who helps create the infrastructure for the production of Wikipedia. And so, our own research was able to piggyback onto the connections that we had in the industry. And so, it is kind of like, it's biased. Like we didn't just have research in the abstract, where we got traction, where we were to, I think, had the greater insights were based upon trusted relationships that we already had going into this research project.

CCB: [00:09:28] Would it? Well, it's a leading question, what might have happened if you had used other organizations as case studies, do you think? If you had time?

Nash Hurley: [00:09:41] Yeah, so we talked about that. The production of Wikipedia, and we try to view it, well we looked at it from the product back to the organization. So, the Wikimedia Foundation is the nonprofit that helps build the infrastructure that allows for Wikipedia to be produced. But we actually did a backwards analysis looking at like, well, how is Wikipedia produced, what organizations, what how does the network work? The other option that we had looked at was potentially looking at a large retail company where there's a physical product. And that has a completely, our hypothesis was that that would have a completely different set of insights that would come out of it from a wholly digital project like the production of Wikipedia. So, you know, the product has a big, the medium of the product has a big influence on what the network is, and then consequently what the workspaces that supports that network.

CCB: [00:10:43] So I want to ask that question about the actual structure of networks, because you spend a lot of time in your research describing them and then ultimately placing, I don't know, the placing of value, but understanding them more effectively.

Josh Emig: [00:11:03] I'm sorry, could you clarify with a specific question was?

CCB: [00:11:08] More the description, if we have folks that are listening that understand some level of network theory, but not to any greater degree? How would you help define the elements of it so that people could start making the link between the theory and the place?

Josh Emig: [00:11:27] Got it. Yeah. So probably the simplest definition of a network is really it's a series of entities or objects or people that have connections among them. And it seems like a very, very simplistic definition. But if you think about, at this point, maybe the most accessible and maybe sort of canonical kind of model of a network might be something like Facebook that most people would really kind of be familiar with. And everybody understands when they're on Facebook that they're connected to certain people and not connected to other people and that their friends are connected to a different set of people than they're connected to. And overall, this structure of connections and what we call "nodes", forms the network. And so it's really sort of taking that idea and thinking about the networks that inform the production of the things, the products that we create and organizations really function as networks. And so that was the sort of jumping off point for us. I think we really sort of landed on a few kind of key aspects. And it's a broad and very, very deep area of study. And there's a lot of people out there that have PhDs in it. So we don't really profess to be able to address it at that level of nuance. But a few of the things that we found particularly important for organizations in the way that workspace impacts organizations are both the structure and strength of the connections, the things or the information that flows across those connections, which I think Nash will probably elaborate on a little bit, as well as the levels of trust and what we would call social capital that is kind of embodied within those networks. All have really big impacts on the outcomes that an organization is able to create.

Nash Hurley: [00:13:44] Yeah, well, you know, I think one of the things that came up to me was an image that we saw over and over again over the last 30 years of research on this topic. And there was just different names applied to it. When it comes to thinking about networks and like a launching off point there are kind of like three general structures of centralized, decentralized, or some sort of mesh network. There are many different names for what those things are, but they're all networks of production. And why I think that that's important, is that in common or like a lot of current language right now, I find that when people talk about networks, they're actually talking about flat organizations. They're talking about not hierarchy, but there's a version of a network that's a centralized network, that's just a hierarchical network. And it has certain properties and certain kinds of cultures and behaviors that go along with it, typically associated with high transaction costs. But then if there's entities like you're, if you're trying to deal with sea level rise, you're concerned about that network dealing with something that has a really high risk, but potentially low transaction costs might take on more like a mesh framework. So I guess the part of it for me is that I think there can be a little bit of a tendency to for the people that are interested in networks to think about flat networks and things that don't have high transaction costs it or high trust networks, to build on what Josh was saying, and I think there's actually a little bit more of a broader perspective that there's like a range of network structures. And it really depends on the cost of transmitting things across the connections. Does that make sense to you, CCB?

CCB: [00:15:44] It does make sense. It does make sense to me. I think the connection that I would like you to make for the Architecture and Design Community, sitting here thinking about this is, what value did you determine understanding this has to developing more effective environments for the organizations that they're supporting?

Nash Hurley: [00:16:14] Yeah, so. You know, Josh, I'll take a pass at this, I mean, I think like a really important thing for me was thinking about the flows that define that end product. So I don't think that, obviously there's always a range of workplace design solutions from architecture down to furniture design. But if you start from the product and you think about the flows that make that product possible, when they're all digital flows, the pressure for colocation is very different from when there's a physical component to the production of that end product. And I think that can come up in surprising ways. A lot of us can look to Facebook, to Google and other companies like that that have a bias towards a digital product and we can look to emulate their workspaces. But I think as soon as there's a physical component in the production of that end product, workspace takes on a very different meaning. So I think that that's like a non-trivial thing of, if you have a digital product or if you have a physical product or physical component to a digital product, looking at examples of workspaces that are designed for digital product production might not be that helpful to you. So I think there's that. And then I think the other part that is there goes into what Josh was talking about in terms of trust and in the qualities of links. None of us make something by ourselves. We always make it as part of some sort of network of production, whether it's five people or five hundred people. But if you're operating in a culture that happens to have a low trust low trust environment, and that just happens sometimes. Once you're there, having a fully distributed organization, it's not going to make that much sense. The most extreme example of this that we came up with just in terms of anecdotally when we're thinking about this was Washington, D.C., There's a value to having physical presence because you're worried about what people are going to say when they're not on camera, or you want to make sure that your ideas are heard or your voice is recognized. And I think that us getting to talking about whether or not people have a low trust or high trust culture earlier in the design process would really help us design better spaces for them. Because you can design a great space, I think, for low trust culture. I think telling everybody that they have a high trust culture and it should be OK with it might lead to some unwanted outcomes. So, sorry, Josh, that them have been too much coffee coming through.

Josh Emig: [00:18:59] I wanted to add another example and I found this, and I recognize this is a difficult subject to try to talk about on a podcast without visuals, because it's a subject that you have to use the language of the networks in order to talk about it. But then you have to define all those terms. So it's difficult. One of the sort of most interesting things I think that was in our Wikipedia case study had to do with the way that they approached what we called convening or gathering. And they have different cadences of where they would bring together, and remember, this is an organization that was distributed prior to the pandemic. And a lot of the patterns that we studied were actually preceded the pandemic as well. And so it just so happened that a lot of other organizations got to experience this firsthand over the last year. But they were doing it prior. And so if you think about, network scientists think a lot about the way that networks affect people's behaviors and the way that information flows in a network. And so there's two things that happen. One is there's a thing called a weak link, which is what you would refer to as an, colloquially probably, as an acquaintance. This is somebody that you're not really in their social group, but you know them. So you can sort of have, there's a communication pathway there, but you're not fully embedded in that group. The opposite of that would be a strong link or a clique, you know, where everybody is very, very tightly oriented. And so, from a network perspective, weak links are very, very good for basically the diffusion of information. So information travels extraordinarily fast through networks that have a lot of weak links. And this is what we call virality, and a lot of things that happen on the Internet happened in this way because things are just shared. When it comes to the adoption of new types of innovations or new types of behaviors, that gets reinforced through cliques and strong links. And so it was really interesting in the case study was that you found that the patterns of convening or gathering that this organization had, basically tried to accomplish these two things. And so you get these very tight groupings where teams are coming together and they're kind of reinforcing their behaviors and they're kind of learning new things. And you have those tight bonds that allow that to happen. And then you have other aspects of these events that are really about, it's like the social hour and there's a very good reason to have a social hour, and that's to create the weak links and diffuse information throughout the organization. And so what was really interesting about that is you had this example of a space, a physical space, playing a very strategic role in the network of bringing people together for the purposes of reinforcement, on the one hand, in terms of strong links and also diffusion on the other in the case of weak links. And I found that particularly informative and salient in the research that we did.

Nash Hurley: [00:22:25] And just to build on what Josh is saying, I think the size of those groups was really, really interesting. Kind of like the 15 person and under for creating those strong links. And so as the group size got larger, the space is changing in the types of links and connection outcomes are different.

CCB: [00:22:47] So it's pretty fascinating, if you think about the results and the data that you've come up with through study and applying that onto some of the actual physical presences that we see in some of these organizations that you're talking about, for example, a Facebook campus that has a football field size, single level environment with lots of perhaps the smaller cliques working together. However, is that intentional or is that serendipitous?

Josh Emig: [00:23:34] I think that it's both. I mean, not to dodge the question, but I mean, I think that we arrive through these things, that we arrive at these things because people do the things that they work. And, you know, there's kind of a whole history of the study of social groupings and sociology. And we sort of tend to kind of arrive back at the same places where we have optimal team sizes of about eight people. And we have Dunbar's number that about a hundred and fifty people seem to be about right. And so I think that, so in a way, it's serendipitous, but it's not because. I mean, in some cases, there might be some intentionality behind it, but the intentionality is just sort of like the optimizing the behaviors of the group over time towards certain outcomes. And I think that what the language of networks does is it gives us a way to talk about it. It gives us a sort of useful abstraction to tame some of the complexity that's involved in those systems.

Nash Hurley: [00:24:40] Well, yeah, I think that one of the things that was interesting to me to think about the overlap between that Facebook example and the Wikipedia production model is the need to continuously own a space. You know, so the production of Wikipedia had a really clear cadence on a quarterly cycle of coming together of their teams to work on the harder problems and build those stronger bonds that Josh was talking about. I think it's interesting in the kind of words coming out this era of what I would frame in my head is like the cult of the large floor plate. The bigger the floor plate, the better would create more opportunity. But I think there's this other component to that, which is like owning your space the whole time or how important it is to have like a permanent spot in that kind of sea of a floor plate versus having a group to go to.

CCB: [00:25:40] So there's also been this past year with all of the upheaval and the immediate dispersal of individuals from location to their own individual spaces, the conversation about is it hub and spoke for the real estate strategy? Does the campus real estate strategy still have any merit? And then the idea of the remote hybrid work environment, which we're probably going to see some of those, perhaps not as much as we thought. And it just keeps changing on a daily basis. So I'm wondering, thinking of the results that you have or the amount of information that you have compiled, where do you see that application in helping build real estate strategies?

Nash Hurley: [00:26:37] I mean. For my part, I do think that, like, I would always go back to the product, I think it's strange that we don't have more differentiation of strategies based upon the nature of the product. Whether it's hub and spoke or fully centralized or fully distributed. I think it comes back to transaction costs. Does space reduce the transaction costs in the production of that product, whether it's a digital product or physical product or combination, as most products are. So I think if you understand the product and you understand the flows that are required to make that product, you understand where there heavy transaction costs. I think at that moment, though, just a few things to put those things in place, then I think you can start to evaluate different real estate strategies. I think it's interesting that we're not hearing the product production first and what kind of space supports that as a secondary thing. And people are looking for a spatial strategy as a lead. I think the other thing that was a really super insightful thing that Ben Waber said in one of our subject matter interviews with him. And Josh, correct me if I'm getting this wrong, but it was this idea that in this last year of more people being distributed and working in a distributed fashion, that some companies might not be aware that they've been spending down their social capital within their organizations, and whether there's different terms for it, whether it's burnout or lack of engagement. But there's a certain repository within the network of social capital that might have been being used up this last year and that needs to be replenished in some way. So one's a question of about flows, the other one's a question of the battery, how much social capital you have embedded in your company's culture, and I think that those are very different workspaces.

Josh Emig: [00:28:36] Yeah, and it kind of hits on, I think a lot of our intention for doing this is that, you know, working in a hybrid way or being and being a distributed organization, there's a lot more to it than just saying that and there's a lot more to it than just saying like, oh, we'll just come to the office three days a week and now we're hybrids. I think the important thing, whether it's like Nash said working from your product backwards and figuring out what works the best or starting with your with your people and the nature of the teams and their connectedness and the information that needs to flow between them and start to optimize for that. I mean, you can come up with a strategy to work in any way. It just so happens that physical space, you know, as we found out, actually does a lot of work in terms of its influence on the network and in the connections among people. But we've also seen that it's pretty important that people have, the work from anywhere type policies, has had an important impact in terms of productivity and engagement and well-being. And they can actually start to drain the social capital battery. But there's actually there's strategies that you can take to get the best of both worlds where you're capturing those benefits while also sort of maintaining those connections and strengthening them. So I think that, yeah, I just think there's more to it than that. And it takes a, it takes a lot of honesty and it takes a hard look at your organization to start to dig into those things. You just have to start asking the right questions.

CCB: [00:30:31] The thing that the I would say the notion that strikes me is how, just to your point, how often are people having these types of conversations? Is it something that we think we will see more of by virtue of giving some form of language around it to the design community? Does that empower them to kind of prompt more of those conversations? It's an interesting piece to think about this research and where it will go.

Josh Emig: [00:31:02] Yeah, I mean, I hope so. I sort of make this joke, but there are different types of cultures that need different types of working arrangements. And so you see a lot of tech companies that have swung heavily toward remote first at this point, and they say they're never going back. We've also seen a lot from the government, finance, and real estate sectors that are gung-ho to get back in the office. And you've also probably seen a lot of stuff in the media or a lot of research out there that indicates that on the whole, that the majority of people are looking for some sort of choice and some sort of flexibility in terms of where they work and when and how they work. And so I think that sort of behooves companies, CEOs, and the people who run these organizations to ask these questions. I think they have to do it if they want to retain the talent that they have.

CCB: [00:32:08] Now. Nash, did you have something else you wanted to say?

Nash Hurley: [00:32:13] No, I think on that topic, in terms of where this goes, I mean, ultimately it comes down to a belief, I believe space is very honest. So a lot of the unproductive reactions that we were seeing workspace design towards the end of the last business cycle in terms of like space is not adapting quick enough to companies evolving needs. I mean, I can't tell you how I heard that from pretty much anybody, any client I would talk to as if it was something new. I think that those kinds of systematic issues are the results of two different networks not being in tune with each other. Building design and building production has always been a process. We refer to it as a product, but it is a process. But it's timestep and it's Hertz regulation, I think is out of step with the speed of growth of a lot of modern businesses. And so I think that that's the kind of like the meta thing, that this research gives people some language around, of getting those two systems to get some standards of interface. And then I think that there's some really honest conversation for leaders in organizations to have realized what kind of culture do they actually have, not the kind of culture they feel like they should have emulating another company, what kind of culture do they have? What is their work product? Do they have to take a physical picture of iPhones? Because that's a physical component. Well, that's going to make sense for people to have a studio to shoot that iPhone in. And so and I think as designers, I think a lot of us have been feeling because know like where's our materials library? I got to get at that. So. Where that's coming to for me is I think if you create the landscape that there's maybe three general forms of networks of centralized, and then decentralized and fully distributed versions, those are all perfectly valid corporate structures based upon, I would say, fundamentally can be understood as transaction costs, and there's totally different workspaces that support each of those types of networks of production. when somebody that has a centralized hierarchical structure feels like their organization should be moved to be more flatter and organizational, flat organization and they ask for that kind of workspace, maybe it's a strategic tool to bring about change. But that friction should be intentional as opposed to a surprising outcome of bringing those two worlds together.

Josh Emig: [00:34:58] I think the other thing that we see happening is that aside from whatever strategy a company might adopt, there's actually, we're starting to build the infrastructure for this now. So you see it like a lot of new companies out there, like Daybase and others that are actually starting to try to provide the means for companies to be more distributed. And I just happen to know them. But there're others that are doing similar things. And so you see co-working popping up in the suburbs and a little bit more far flung places. And so, again, I think that a lot of these things will kind of evolve and parallel that companies will be able to evolve their thinking about how their people work best. And we're also seeing a lot of more opportunity. Whereas if you go back to what Nash termed, and I think the dominance of the large, centralized floor plate. You know, a few years ago you had WeWork in a lot of kind of localized co-working companies, but the opportunities weren't really there. And so now you see real estate starting to evolve in this direction as well. And I think it's going to play out over a longer period of time. I don't think that we're going to snap into anything that's decisive by the third quarter of this year, it's going to take a little longer than that.

Nash Hurley: [00:36:32] The build I would have on what Josh was saying, I think is an interesting reference from 1930 if that's right, Josh, the Coase's "Theory of the Firm"? But it looks at transaction costs and why firms exist. And at the time it was explaining a lot of manufacturing tendencies of, when does it makes sense to have manufacturing under one roof? And when does it make sense to buy things from the market? And I think what's going on in the wake of let's call it Red One of the WeWork era, and there's the new more evolved offerings that are coming on is the marketplace is advancing. It's more competitive. So what you can get from the market versus what you have to hold in house through your own facilities team is really different. It is not 10 years ago and it's certainly not 20 years ago. And so if I can go out into the marketplace and get a workspace, I now have options. I don't have to go to the manufacturing. I don't have to have the forge inside my place. I can just buy a finished part. And I think that's kind of exciting for me. Looking to the future and what it means for workplace designers is that maybe the thing that we're designing and working on is not the widget you can go by out there in the marketplace, it's something that's really specialized in particular to the organization. Something that is a less commodified version of space.

CCB: [00:38:02] Wow. OK, Josh and Nash, you have given us an enormous amount to think about, and I'm going to also applaud your ability to describe things that have so much more theory attached to them in a fairly concrete way so that I think most people will be able to understand. Thank you again for your time. We are so grateful that you participated in The ONEder Grant, and I'm going to say goodbye.