4. The Death of Distance?: Measuring the Power of Proximity

With companies not just to a single location, but more and more frequently spread across the globe, how much does having a water cooler (see Chapter 3) really matter today? It’s not hard to imagine that the water cooler effect will gradually fade in importance as workforces become more and more distributed and as technologies allow workers to collaborate more effectively across huge distances.

The revolution in communication technologies continues, slowly lowering the barriers to remote work. From e-mail to instant messaging to video conferencing, each has served to make collaboration over distance much easier than in the past.

For decades these communication technologies did not fundamentally change the way that people worked. Yes, now companies could have branches in different locations, and as an employee, you might be working with someone whom you didn’t meet face to face, but you were still going to work in an office. You still commuted to work and had coworkers who were physically in the same place as you. In the 1970s, however, this model started to shift.

The term telecommuting was first coined in 1973 and encapsulated this change in work. Information technology had advanced to a point where some people could do their job from home part of the time. Particularly for people working in computer-intensive jobs such as software development or even factory management positions, telecommuting can be an attractive option.

One of the main benefits of telecommuting is that workers are able to balance their personal and professional responsibilities. If your child is sick and has to stay home from school, for example, telecommuting gives you the ability to work from home and take care of your child at the same time. As any working parent can attest, this benefit offers substantial relief from the frantic scrambling to find a babysitter that typically occurs in this situation.

Telecommuting was originally conceived as an occasional perk. If sometimes you needed to work from home, you could feel free to work a few days a month in your pajamas. Over the following decades, however, this model also started to change. Instead of people staying at home one day a week, this gradually stretched to two days, then three. Eventually, some organizations found it desirable to have people telecommute every day.

One of the places where telecommuting first manifested itself was call centers. In call centers, having employees talk to each other was never thought to be important, so having people work from home was a logical step. Physical space could be reserved for someone else.

Telecommuting also began to expand over time. Rather than having individual workers opt into these programs, companies began closing down entire offices and making whole branches of the company virtual. Now some companies just need to pay for an employee’s Internet connection and office supplies. There are no leases to sign, no cleaning crews to hire. The reduction in employee commuting times also meant that people could spend more time with their family and even more time working if they so chose.

This change hasn’t been confined to call centers. Organizations in nearly every sector are embracing what are being called “virtual organizations.” Two guiding principles of these organizations are efficiency and flexibility. The idea is that these organizations are efficient because moving physical bodies around wastes a lot of stuff: time, fossil fuels, and the monetary cost of maintaining a physical office. The average American spends about 50 minutes per day commuting. Splitting that time evenly between family and work would enable people to spend 5% more time working—not an insignificant number.

These organizations can also be more flexible because telecommuting employees are able to choose whom they spend time with. Because they can’t bump into people in the hallway or be roped into lunch with coworkers, they can’t be randomly “interrupted.” Instead they have to rely on e-mail, the phone, and tools such as Skype to talk to other people. All of these tools are used in a completely intentional way, so telecommuters can set up meetings and calls only with those people with whom they absolutely need to communicate.

An interesting parallel to the telecommuting trend is offshoring. Offshoring generally involves taking a part of a company and placing it in another country, typically one where the cost of living is far lower than in the original location. Classic examples of this trend include transplanting call centers from the U.S. to India or moving software development to eastern Europe. Although these offshored divisions usually need to interact quite frequently with their onshore counterparts, the rise of remote collaboration technology has made offshoring a much more palatable option for business leaders.

The decision to offshore an operation is nearly always based on economic considerations. As education levels in the developing world have caught up to more traditionally advanced countries, there is no longer a strong concern about a drop in production quality due to unskilled workers. In fact, because people in developing countries often have to deal creatively with challenging problems, their location can actually contribute to greater innovation.

An intriguing example of this phenomenon comes from the Nirma Institute of Technology in India. In the early 2000s, students in the computer science department of that university wanted to do research on virtual reality. Unfortunately, even the least expensive virtual reality gloves cost hundreds of dollars. This was well out of reach for the vast majority of students at the time. They had to develop a lower cost alternative to perform their work.

Playing around with a computer mouse, they hit upon the idea of attaching multiple mice together into a glove-like contraption. A string would attach to each finger, which would then attach to a mouse wheel. Bending the fingers in a particular direction would cause a corresponding movement of the mouse wheel, which would be fed to the computer. In that way the students were able to create an extremely accurate substitute for a virtual reality glove for around $20.

Companies look at examples of this kind of ingenuity and realize that if they can get one of these students to work for their company at half the cost, they are probably getting a good deal. As this trend accelerates, suddenly companies are no longer considering offshoring one part of an organization but many. Most major organizations have operations on multiple continents and dozens of countries. IBM, for example, operates in more than 170 countries. Given that as of 2012 approximately 195 countries exist, IBM is about as “global” an organization as you can get. Most companies that make extensive use of remote workers and offshoring, however, don’t look at the results of these changes beyond the immediate impact on their personnel costs. In order to determine the real effect of these practices, you to need to look at some hard data.

First, let’s talk about telecommuting. It represents a flexible kind of workplace, where you only stay home when going to work is very inconvenient. In general, few downsides exist for this arrangement. Sure, you might miss a few interactions at work, but from that perspective, a day at home telecommuting is essentially equivalent to taking one day of vacation. Ample evidence shows that vacation and recharging in general are good for worker performance, and if you had to go to work knowing you had five hours of errands ahead of you when you got home, you probably wouldn’t be very effective.

Ideally, however, telecommuting should be used somewhat infrequently. The principle behind telecommuting is that you can work from home when necessary, but you’re still coming into the workplace the vast majority of the time. By contrast, when people only work from home, they are completely cut off from the physical workplace.

Let’s consider the case where you’re the only person working from home while the rest of your colleagues continue to go to the office. The time your colleagues would have spent interacting with you face to face is reallocated to

• Interacting with other people

• Communicating with you via other channels

The main interactions that are lost are social in nature: the bumping into other people and the schmoozing that happens before and after meetings. Because they’re not strictly work related, these interactions are much less likely to be continued after someone becomes a remote worker. Communication that is necessary to complete specific tasks, however, shifts to e-mail, phone, and video chat.

So how does telecommuting impact individual performance? Let’s look at the numbers. In the call center study from Chapter 3, my team and I were able to measure the effects of face-to-face interaction, and we also collected e-mail data. E-mail was a perfect representation of the org chart. In other words, peers didn’t communicate with each other electronically, and e-mail communication in general had no correlation with performance. Removing face-to-face interaction and cohesion from consideration in our model of productivity leaves an average call handle time of 297 seconds.

The average call handle time in the original dataset was 263 seconds, which implies that if we completely eliminated face-to-face interaction for an individual call center employee, we would expect performance to decrease by 12.9%. This is an incredibly significant effect. If we sent all call center employees home, we would have to hire 12.9% more people to handle the same call volume. At an average salary in the U.S. of around $30,000, in a call center operation such as Bank of America’s with more than 10,000 call center employees, this translates into an additional expense of at least $38 million.

This equation also leaves out the mental toll this action would take on workers. We would expect to see stress increase quickly as well, with a 13.1% increase in stress if we remove face-to-face interaction. Not only would removing this interaction make a hard job even harder, but it would contribute to turnover and further add to the cost of running a call center operation.

These costs have been examined in creative industries as well. For better or for worse, some teams are required to be entirely virtual. If you’re trying to run an international advertising campaign, for example, you need people on the ground in many different countries collaborating with each other. In general, these teams are far less effective than co-located teams (teams in the same location). They trust each other less, and completing tasks takes them longer.

Elena Rocco from the University of Michigan designed an ingenious study¹ to test some countermeasures for low distributed team performance. Rather than only study whether face-to-face groups work better than remote groups, Rocco had one experimental condition where people met face to face before going off and working remotely. The idea was that by getting a feel for the other people on your team when you met in person, you would know how to respond to them better and interact on a deeper level when you were no longer co-located.

In her study, face-to-face teams were the highest performers, but they were followed quite closely by teams that met face-to-face first and then worked remotely. Groups that only worked remotely were by far the lowest performers. This study shows that getting everyone together in person before kicking off a project is probably worth the cost of a few plane tickets, even though doing so is not practical in all circumstances.

Offshoring is a different beast altogether. When a company places parts of the organization in separate locations, collaboration within a division is not necessarily the issue. Rather, problems come from coordination between divisions. Language problems can, of course, contribute to these difficulties. At a basic level, if people literally can’t speak the same language, then they will find collaboration challenging. Beyond that issue is the potentially acrimonious relationship that can develop between groups who are based in different locations.

Michael O’Leary and Mark Mortensen from MIT explored this problem by examining the configuration of groups that were collaborating in different locations.² Specifically, they looked at the number of team members in different sites and saw how this distribution related to team success. Their main hypothesis was that an imbalance in the number of people at different sites could cause individuals to identify as part of different groups, drastically reducing trust and productivity.

One might believe that the evenly balanced or far out-of-balance teams would have the most trouble. After all, two groups of equal size might both want to claim ownership of a project. On the flip side, if one group is vastly bigger than another, you would expect them to run roughshod over the smaller team. What the study found was a different story.

Groups that were relatively even were actually quite productive. Because the teams in different locations were of equal size, they didn’t feel threatened by the other group. This enabled them to work out their differences equitably. In the very imbalanced groups, a similar effect was observed. In this case it was never a question of what group was in charge, because the larger group could always override the smaller one. With that issue settled, conflict around project control faded into the background.

The worst teams had one group that was slightly larger than the other group. The minority party knew they were outnumbered, but resented the fact that they could be excluded from the decision process by the larger group. This often degenerated into pointless bickering, which had strong negative effects on performance.

Notice that this discussion doesn’t even consider organizational structure. This is simply the number of people in different locations. The takeaway is that beyond the necessities of setting up meetings and creating an org chart that effectively ties in remote workplaces, social considerations have to be accounted for as well. You need to think about how to set up these remote workplaces in such a way that people feel like they’re collaborating with each other rather than working at odds.

As discussed earlier in this chapter in the case of creating a virtual organization, meeting face to face before starting work on a project is critical. The same can be said of offshoring. If you want things to go well, if you want people to care about each other and think of themselves as part of the same team, you need at least some team members to meet face to face.

How do companies get the right people to talk to each other, or even get people to talk in the first place? Instead of focusing on using formal meetings to this end, this section looks at how to organize the layout of a workplace to make it more likely that the right people will talk to each other.

It should come as no surprise that you’re not equally likely to talk to everyone in the office. After all, some office buildings house tens of thousands of employees. In these skyscrapers, companies can span dozens of floors that themselves can be thousands of square feet. As a matter of common sense, you are not likely to talk to someone 20 floors down from you. Conversely, you should be extremely likely to talk to someone who sits next to you.

Expanding on this theory a bit, the general idea is that the probability that two people will interact is inversely proportional to the distance between their desks. Because you can’t walk in a straight line from one desk to another without having to jump over cubicle walls (or through them), the theory takes into account the distance you have to walk from one desk to another. To calculate this distance for our study, we drew lines across corridors and hallways, paying attention to the points of intersection between these different lines. These maps allowed us to automatically calculate the distance between people who sit on the same floor.

Dealing with people on different floors of the same building might seem simple at first. You could just add the vertical distance you need to travel on the elevator to get to the other person’s floor. This distance, however, doesn’t capture the actual amount of time needed to travel from one floor to another. If you take the elevator, you need to press a button, wait for the elevator to come, get into the elevator, press another button, wait for other people’s stops, and then finally get off on the right floor. Although people can take the stairs, the physical effort required often dissuades people from taking them even for a single floor.

These points lead to the conclusion that being on a separate floor is fundamentally different from being on the same floor. Even the difference of a few staircases should have a powerful effect on interaction patterns, one that’s slightly different from the concept of distance.

This makes the company campus setup all the more interesting. Company campuses are made up of a number of buildings housing different parts of the organization, all grouped together in a single location. Some campuses can sprawl for miles, whereas others can be smaller and encompass only a few clustered buildings.

Campuses are particularly hot in technology companies, with juggernauts such as Google and Facebook buying hundreds of acres of land to build, in essence, a company city. Google’s campus, for example, is replete with beach volleyball courts, bowling alleys, fitness centers, and dozens of cafes and restaurants that serve Googlers some of the finest “cafeteria” food in the world.

These campuses create a strong sense of community. People throughout Google and Facebook eat the same food, go to the same gyms, and play the same games. Beyond being fabulous perks, these shared experiences make relating to each other easier for people in different parts of the organization.

The question is: Do these campuses actually yield better interaction patterns? One of the major reasons companies invest in these campuses is that it’s believed they promote closer collaboration between different parts of the company. Although the possibility for interacting face to face technically exists when you’re on the same campus, whether people can make use of this co-location to do more than simply organize meetings is an open-ended question.

An issue is that walking between buildings takes an order of magnitude more time than going between floors. You normally have to take two elevators or two sets of stairs in addition to walking across the campus to get to your destination. As such, it is incredibly unlikely that this will lead to serendipitous conversations. Even without looking at the data, one could be quite confident that people in different buildings on campus won’t hang out in one another’s coffee areas.

That’s not to say that serendipity can’t happen in company campuses. At company-wide talks or interest group meetings, bumping into someone who sends your thinking in a completely new direction and starts off a new collaboration is entirely possible. Indeed, this serendipity is probably the biggest benefit of company campuses. However, serendipitous interactions are probably not going to happen in the spaces between buildings because you walk with people you already know. This shifts the focus in company campuses from the actual office buildings to the event spaces and central social areas where people from other areas can actually meet.

Companies such as Google get this concept. The company offers a variety of classes and speaking events from outside researchers and personalities that pique the interest of different constituencies at Google, creating many opportunities for division-spanning collaboration to occur. Companies need to take these events into consideration when planning their new campuses. Otherwise, they risk creating a bunch of corporate islands that happen to be in the same pond.

As employees can move around within these campuses and the workforce becomes more mobile, even the one desk per person standard has shifted. The concept of open seating in particular has gained in popularity over the last decade, enabling people to choose their own desk on a daily basis within their team’s area. One of the main benefits of such areas is that as collaboration needs change, people are free to rearrange their seats to sit closer to the people they need to talk to.

ESPN is one company that has embraced this model for its video editors. For those who haven’t watched ESPN, they run live sports broadcasts and slick analysis and news shows that seamlessly intersperse commentary and event clips. They are a massively successful organization, holding a nearly insurmountable share of the U.S. sports market and commanding huge premiums in terms of cable fees and advertising prices.

While they have huge resources at their disposal, what always interested me about this organization was that they would have clips of some of the most obscure sporting events going on in other parts of the world. When I visited ESPN’s headquarters, for example, I saw clips of professional Polish volleyball, African cricket matches, and minor league baseball games. When an interesting play happens in one of these matches, it’s not unusual to see it fed to a number of the network’s programs to be disseminated to the wider world.

I had always assumed ESPN had people watching major sporting events in the U.S., but I also assumed they were alerted to relevant plays in obscure sports by media partners in other countries. Wrong. ESPN has dozens of video editors, watching every single sporting event that is broadcast anywhere in the world every hour of every day. This means that if you walk into the editing room at 3 a.m. on a Tuesday, you’ll find a bunch of editors sitting at their desks watching sports.

The reason that ESPN uses open seating becomes much clearer when taking into account this work style. If many sporting events are going on, then more people need to be sitting around watching sports. For major events, usually multiple people are watching from different angles. Having these people sit near each other is better so they can discuss how interesting a particular play is. Assigning people set desks at different time periods could probably reduce the overall number of seats, but the room would become significantly overcrowded during huge sporting events such as the Olympics or March Madness.

ESPN has even extended this model to their individual television programs. Each team responsible for a television program sits in an assigned area, a small part of a larger bullpen where the producers and editors for the shows are located. The desk areas rotate over time, so that the morning Sportscenter crew swaps desks with the 6 p.m. Sportscenter staff in the afternoon. Within each seating area, people can just log in to a computer and sit down wherever they want. This capability is particularly important when they need to coordinate with people on other shows. You want to check how the shows before and after you are planning on covering an event so that you make sure storylines aren’t repeated ad nauseum and to facilitate unique takes on these events with your colleagues.

Intuition suggests that distance matters, and some of the world’s most successful companies are embracing this fact. To discover the levers that really make these distance mechanisms work, however, we need to look at the data.

There are many ways to look at distance data, but this discussion simplifies things a bit so that the numbers are easier to understand. In particular, our studies take into account four different distances between people:

• Next to each other

• Same row or hallway

• Same floor (or less than 50 meters apart)

• Different floor (or more than 50 meters apart)

These distances can, of course, be differentiated by counting the number of steps between desks and looking at big scatter plots; but what you’ll see is that after you get past the same hallway, the effective probability of interaction drops precipitously.

The following discussion looks at two badge datasets, one from a German bank and another from an IT firm. In these studies we knew exactly where people sat, and for a four-week period we knew who spoke with whom. The next step is to see what fraction of these interactions occurred between people at different distances. Let’s go over a simple example.

In Table 4.1, each entry represents how much one person spoke to another. For example, the entry in row B, column A indicates that B spoke with A nine times. In this example, let’s say that A and B sit next to each other (distance 1), but C sits in another row (distance 2). Let’s group these interactions by distance.

As you can see from Table 4.2, in this example people who sit next to each other always talk to each other, but only 1 interaction occurs at distance 2.

In the German bank study data, we also looked at e-mail communication as it relates to distance. This issue is interesting because e-mail is not bound at all by physical constraints. E-mailing someone sitting next to you takes just as much effort as e-mailing someone on the other side of the world. If e-mail also falls prey to distance effects, however, then the implication is that to spur truly global communication, companies need to rely on other communication channels as well.

Some early work on e-mail communication and distance effects comes out of the University of Toronto. In their study, people in a pharmaceutical company were randomly assigned offices when they joined the company. As a result, people in the same workgroup were now no closer than people from different workgroups. If distance influenced communication probabilities, then it must be due to distance effects rather than formal work needs.

As you might suspect, they found exactly what we are hypothesizing that we’ll see with face-to-face communication: the greater the distance between people’s desks, the less likely they were to communicate over e-mail.

Now you have a pretty good idea about what we expect to see when we look at this study data. So let’s get right to it.

Putting the data from both companies on the same graph reveals something interesting: the negative relationship between distance and communication is nearly the same for both organizations, even when we throw e-mail into the mix. These are two very different companies. One has individual engineers in the midwestern United States work on individual tasks, while the other has people working collaboratively on different aspects of financial products in Germany. They have employees from different cultural and educational backgrounds, different organizational goals and formal processes, and yet the distance between desks appears to be one of the main drivers of interaction.

Getting into the specifics, the downward trend observed as distance increases is not surprising given the previous discussion. The likelihood that you’ll bump into somebody, the chance that you’ll make the extra effort to specifically swing by someone’s desk, gets drastically lower as distance between desks increases.

The decrease in e-mail communication, while in line with previous research, is still incredibly fascinating. What makes e-mailing people in different locations so difficult?

To some extent, the difficulty of cold calling explains this phenomenon. Although some naturally outgoing people don’t have trouble calling up someone they don’t know to ask a question, for many people this task is quite difficult. E-mailing someone you don’t know can be just as awkward.

The German bank data allows us to test this theory. With the data, we’re able to actually observe how much you e-mail people you spend time with versus people you don’t spend time with and detect any differences. If our hypothesis holds, then the amount you e-mail someone should be strongly correlated with how much time you spend with that person.

An analysis of the German bank data provides results that at first seem surprising. In general, the amount an employee e-mails another person is not related at all to how much she talks to him face to face. Remember from the overall relationship between distance and communication probability, however, that the farther away people’s desks are from each other, the less likely the two people are to communicate. Taken together, these results imply that people switch off using face-to-face and e-mail to fulfill their communication needs.

Think about this result with a simple example. If I talk to someone a lot face to face, it has no bearing on how much I e-mail that other person. But suppose I need to communicate 10 “bits” of information to that person. Let’s say that one face-to-face interaction can transmit 2 bits of information, while e-mail can transmit 1 bit of information. There are many ways that I can get to 10 bits: 5 face-to-face interactions, 10 e-mails, 3 face-to-face interactions and 4 e-mails, and so on.

Now assume that for different people I need a different number of bits. So for Bob I might need 10 bits, whereas for Alice I might need 15 bits. I could choose to mix these bits in different ways, with some bits being filled more by face-to-face communication but other bits being filled more by e-mail communication.

Drilling into the data a bit further provides some insight into how people might make these decisions. For one class of relationships—people who sat next to each other—the study results showed a correlation between e-mail and face-to-face communication. In that case, the amount of e-mail was strongly negatively related to the amount of face-to-face interaction.

This makes a lot of sense because if you sit next to someone, just leaning over and asking him a quick question is easier than sending him an e-mail. As someone gets farther away from your desk, however, the type of information you want to communicate likely plays a role in what communication medium you choose. As discussed in Chapter 3, e-mail is excellent at disseminating easily codifiable information, whereas face-to-face communication is effective for communicating nuanced and complex information.

Long desks increase distance between people, and thus should have a chilling effect on the amount of interaction. Private offices increase distances even further. Smaller cubicles are positive from a distance perspective, but high cubicle walls could potentially make communication more difficult.

The question of desk size in particular is interesting because it seems like such a minor difference. After all, a six-foot-long desk versus a seven-foot-long desk doesn’t seem like a big deal.

In our Bank of America project, we were fortunate to have groups that sat at desks of different lengths. One group had longer desks with low cubicle walls, while the others had traditional cubicles with smaller desks. The group with the longer desks had 43% less intragroup communication, while the groups with cubicles were internally very tightly connected.

These effects aren’t only limited to desk size. Other seating areas around the office are also strongly impacted by distance. In a study at an online travel company, the size of the lunch tables was a strong predictor of future communication and individual performance.

At this company, people could eat lunch at their desk, in a small café area, or downstairs in the main cafeteria. Obviously eating lunch at your desk meant eating by yourself, while the small café area had tables that could accommodate up to 4 people. The cafeteria, on the other hand, had tables that could fit up to 12 people.

After collecting badge and e-mail data for four weeks, we saw that, not surprisingly, after an employee ate lunch with someone, he was more likely to talk to that person during the rest of the day. Eating lunch together enables people to get a sense for what everyone is working on and what problems they’re having. This deep knowledge is crucial if you ever have an issue with your own work because you’ll have a strong network spread across a diverse group of areas, making it more likely that you’ll know someone who can help you find a solution.

What was interesting, however, was that people who ate lunch in larger groups weren’t going to the cafeteria with that entire group. The larger tables forced different groups to sit together, which caused smaller lunch groups to merge into a “super-group.” This activity wasn’t possible in the café, because the tables were too small.

These larger lunch group interactions directly led to higher group cohesion and consequently to higher performance. People who interacted in these large lunch groups were 36% more likely to interact with each other at some other point in the day compared to other groups, and they were also significantly more resistant to the effects of stressful events such as layoffs.

At the root of these incredible effects, however, was a simple element: a longer lunch table. The decision to buy those tables had a significant impact on this company and its employees. These results led one of my colleagues to suggest: “Maybe we should just duct-tape smaller tables together!”

That assumes, however, that you just want to talk to a lot of people, rather than the right people. What you should be trying to figure out is whom you should be sitting close to, both for getting your job done and for making good connections.

Companies also need to think about the importance of office layout. Rather than sticking people into open desks just because they’re open, managers need to consider the effects on collaboration and attempt to create a layout that makes it easy for the right groups to communicate with each other.

Beyond the layout of the space, furniture choices have effects as well. The size of your desk impacts not just how likely you are to talk to the people sitting next to you, but by extension the probability that you’ll interact with people all across the office. Taken to the extreme, this means we should all sit in single chairs next to our colleagues to minimize the distance between us. This underscores the tension between the need to have a desk to get stuff done but also to create an environment where you’re closer to the right people. Clearly, employees need some kind of desk in order to work, but maybe it’s not the sprawling workspace in a private office that has been an ideal. Figuring out the right balance between desk space and communication requirements is crucial to creating a workplace that fits the needs of employees and the organization.

Office furniture also includes common seating areas and cafes, which have become staples in the modern company. As our results at the online travel company show, the choice of a small, round coffee table versus a large, square lunch table takes on an added, serious dimension. These areas are foci of social activity, and ensuring that they not only look nice, but also have real, tangible benefits for the company is key.

Although the study found that larger lunch tables made people more productive, that will not necessarily be the case in other companies. When sitting down for lunch, you need to ask yourself: What kind of connections do I need to do my job, and what pattern of collaboration will be most supportive for me right now? Rather than sitting with your best friends, maybe branching out one week to sit with a larger group from another department makes sense. This would make talking with those people later in the week easier and would provide a window into other activities around the workplace.

Although in this chapter we’ve discussed increasing the likelihood of interaction, we can also use this effect to decrease communication between different groups when it’s desirable. For example, the accounting division might not need to communicate much with the research division, so managers should try to put more space between these divisions to reduce extraneous communication. Because a workday contains a finite amount of time, spending time wisely and making sure the environment is set up in such a way that employees will naturally talk to the right people are important.

Distance is a natural lever that companies can pull to get predictable responses in collaboration patterns, but it doesn’t force people to behave in a specific way. If I want to talk to someone on another floor, I’m perfectly capable of taking the elevator and walking to his or her desk. However, when I’m simply walking around the office and not looking to talk to anyone in particular, I’m much more likely to bump into the people who are close to me. The role of individuals and company leaders is to engineer the environment so that those serendipitous interactions are as useful as possible.

This chapter covered a lot about remote work and telecommuting. A large, robust body of research indicates that companies need to figure out some way to get employees together face to face, because remote work tools just aren’t cutting it. The more that employees can meet in person with their team, the more effective they will be and the better the team will perform. When making the decision to work from home, employees must weigh personal needs against the negative impact it will have on their colleagues and on them as individuals.

Although remote work is certainly a reality and will have to continue into the future, people need to pay a lot more attention to communication to make those arrangements work. Face-to-face meetings are an investment, not a cost; the small amount of resources that are put into physical space comes back many-fold in increased productivity and job satisfaction.

Companies should always look to physical space as a key part of their toolbox for changing patterns of collaboration and behavior. The actual layout of the office, the type of furniture, and the decision to let employees work remotely all have a profound impact on both companies’ and individuals’ success. Distance is not dead. If anything, it’s more central to our lives than ever.