Keywords

1 Introduction

As work teams become more distributed, and the Internet is more widely accessible, collaboration in the enterprise no longer happens exclusively in face-to-face meetings. Employees are commonly found to be working from diverse locations with non-traditional schedules. Additional classes of workers have emerged – from the teleworker who works from home one or more days per week to the mobile worker who works from coffee shops, airline terminals, and customer sites. Online communication and collaboration tools have greatly enhanced employee mobility by allowing them to stay in constant contact with their colleagues, but what can we learn from how teleworkers use these tools? In this case study, we delve into one company’s teleworker population and how its usage of enterprise social media differs from populations that work from standard office locations.

2 Background

The MITRE Corporation is a not-for-profit organization with expertise in systems engineering, information technology, operational concepts, and enterprise modernization. In addition to managing multiple Federally Funded Research and Development Centers (FFRDCs), MITRE supports its own independent technology research and application development for solving sponsors’ near-term and future problems. MITRE has over seven thousand scientists, engineers, and support specialists who work on hundreds of different projects across various sponsors and numerous domains. MITRE has two principal locations – one in McLean, VA and one in Bedford, MA – along with additional sites across the country and around the world. Over the past decade or so, MITRE has seen a shift of its working population from campus-based and site-based staff to mobile populations (those with no permanent, dedicated office space), telecommuters (those working from home one or more days per week), and teleworkers (those working from home full time).

Because of the high level of technical, operational, and domain knowledge required, staff often seek out and consult with other MITRE experts on particular topics or seek assistance from in-house librarians in gathering resources. Much of MITRE’s work involves collaboration across time and space, requiring virtual and mobile teams to share these resources and relevant research internally, but also to collaborate and share with external partners including government agencies, industry, academia, vendors, and other FFRDCs. Business and social connections formed during the duration of specific projects and programs often become hard to track and manage over time, and there is concern that knowledge may be lost between engagements.

To solve some of these problems and pressing business needs, the MITRE Corporation pioneered its exploration of enterprise social media in 2005 with research into social bookmarking tools, which were rapidly growing in popularity on the Internet at that time. MITRE developed and piloted its own internal version of a social bookmarking tool for resource management, information sharing and discovery, expert finding, and social networking [1]. Two years later, MITRE began to look at social networking tools as a means to track external relationships and to collaborate with external partners. In 2009, MITRE customized and launched Handshake [2], a social networking platform based on the open source Elgg platform [3]. The tool was made accessible outside of MITRE’s network, so that the company could connect and engage with external partners while simultaneously enabling employees to collaborate with each other internally.

Handshake was initially released to just a small community of early adopters, but its user base quickly grew by word of mouth. Over the past five years, over 13500 members have joined Handshake, including 4000 invited external participants. Members have created more than 1500 Handshake project spaces, communities of practice, peer support groups, social groups, and more. Handshake enables users to participate in discussion forums, share documents, collaborate on wiki pages, and “connect” (or “friend”) each other to keep abreast of what is happening in their network. Handshake also provides activity streams both within and outside the tool along with customizable email notifications to alert users to new content, comments, and updates.

3 Related Research

Much research has been done on the teleworker experience: the communication challenges imposed by the lack of common ground [4], factors like geographic separation [5], and even respect among peers [6]. Better collaboration tools make the mechanics of communication easier, but, as Olson et al. have put it, “Distance still matters” [7].

Despite the challenges, teleworking continues to appeal to employees seeking to improve their work/life balance, eliminate commuting hassle and expenses, and reduce stress. An analysis of existing research performed by Gajendran and Harrison [8] declared that the net effect on the employee is positive:

“Telecommuting has a clear upside: small but favorable effects on perceived autonomy, work–family conflict, job satisfaction, performance, turnover intent, and stress. Contrary to expectations in both academic and practitioner literatures, telecommuting also has no straightforward, damaging effects on the quality of workplace relationships or perceived career prospects. However, there is a downside of higher intensity telecommuting in that it does seem to send coworker (but not supervisor) relationships in a harmful direction.”

It is no surprise, then, that more people are choosing to work remotely: Forrester Research predicts that the ranks of telecommuters will swell to 63 million by 2016 in the US alone, with 11.7 million working from home full-time [9]. Today’s teleworkers benefit from years of accumulated knowledge about working remotely. The United States Government even hosts a purpose-built web site, telework.gov, to provide tips to teleworkers and their managers.

A lot of theoretical knowledge about telework comes from studies of teams that use Information and Communication Technology (ICT) to collaborate over distance - what Olson et al. described as “hub to hub” communication [10] (where groups of people in two different locations collaborate with each other across distance). Each cycle of the ICT evolution, from email to video conferencing to instant messaging, is accompanied by research into the effect of that technology on remote collaboration. Enterprise social networking, in turn, has been explored by early adopters [11] as well as researchers surveying a developing technological landscape [12].

More recently, researchers have started taking a more nuanced look at the teleworker experience, which traditionally follows the “hub and spoke” arrangement (where the teleworker is removed from the collocated team) [13].

This paper connects the two threads of research (enterprise social networking with teleworker experience) by evaluating the specific case of Handshake adoption and usage by MITRE’s teleworker employees.

4 Methods

The Handshake platform was instrumented to collect usage data and statistics, which provided us with a rich set of data to analyze. For this study, we looked at tool adoption rates; login patterns and frequencies; contributions in the form of posts, comments (directly via browser and indirectly via reply-by-email), and edits – by volume, frequency, and pattern (i.e., time of day and changes in volume or frequency over time); number and type of group memberships; groups created and (co-) owned; and number of connections. We examined the data from the perspective of various user demographics including job level, hire date, work unit (organization), time zone, location, and residency status (e.g., resident on campus, site-based, mobile, or teleworker). By looking at the data across the different slices, we were able to detect patterns and differences, most notably by location and residency status. Due to the fact that the data lacked a normal distribution, we used non-parametric statistics such as Chi Square and Mann-Whitney U to analyze the data.

To complement the quantitative usage data, we reviewed and analyzed responses to a user survey first administered in September 2011 (333 respondents) and later repeated in November 2012 (409 respondents). Survey questions focused on aspects of usability, business value, knowledge management, situation awareness, collaboration support, and social connections. The survey snapshots represented user satisfaction over time for the general Handshake population. We were also able to take a closer look at changing satisfaction ratings for 99 users who participated in both surveys; we analyzed their rating changes with longitudinal changes in their usage patterns. Again, in the survey data, we were able to detect differences across the different populations. Respondents to both surveys were encouraged to write in comments, which helped us to better understand their responses and allow us to craft follow-up questions.

Based on our observations and analysis of usage patterns, survey data, and user comments, we developed several hypotheses about the differences we saw between the teleworker population and campus and site residents. We then developed a set of questions designed to probe the targeted group of teleworkers through structured interviews. We solicited volunteers who were willing to participate in a half-hour interview over an audio connection. Of those who responded, we selected 21, making sure that they represented all company centers and job levels and included new hires, more established employees, and long-term employees (over 15 years at the company). We also spoke to participants who were hired directly as teleworkers as well as those who transitioned to teleworking status after working on campus or on site. The interviews were transcribed and the data was later tabulated and analyzed.

5 Results

We analyzed both quantitative and qualitative data to understand our user population. In the course of our study, we observed the following differences between teleworkers and their office-based colleagues, based on our quantitative data:

  • Teleworkers are early adopters: In our analysis of the data, we found that teleworkers have the oldest account age, i.e., they became members of Handshake earlier than other populations. Their median account age is ~130 days older than the campus-based populations, and the mean is ~160 days more than other locations (p < 0.001). When Handshake was initially launched, there was also a higher percentage of teleworkers who were members; 65 % of the teleworker population joined within the first year, compared to only 40 % of the rest of the company.

  • A higher percentage of teleworkers are contributors: ~42 % of populations from principal locations contribute content and post comments while 35 % contribute from sites. In contrast, a whopping 61 % of teleworkers are contributors, a number significantly higher than other employee populations (p < 0.001).

  • Teleworkers are at least on par with other locations in all measures: In all the data we analyzed, we did not find that teleworkers have lower usage statistics than any other population in the organization. In all measures (e.g., average number of groups, average number of connections, percentage of contributing population, average number of monthly contributions, etc.), they are either the same or higher. This speaks to the level of engagement of teleworkers; we found teleworkers to be either as active as or more active than other populations. The notable exception is that certain types of groups are not as popular with the teleworker population (see below); however, teleworkers do not join fewer groups than other populations, in general.

  • Teleworkers interact across more diverse hours: Staff in main offices tend to work a conventional 8am-5pm schedule and post or interact on Handshake infrequently during after hours. Comparatively, teleworkers have a far more diverse interaction pattern even when accounting for time zones; they post to Handshake both earlier and later in the day, indicating they that are online for an extended work day. This may speak to the lesser boundary between work and home life that teleworkers talked about during our interviews.

  • Teleworkers are less likely to use networking tools for social uses: We found that teleworkers are members of work-related groups more often than social groups. Other populations belong to a higher percentage of social groups (p < 0.01).

6 Discussion

There were a number of ways in which teleworkers engaged with our enterprise networking platform differently than employees at MITRE campuses. First, we discovered that teleworkers are earlier adopters of the enterprise networking platform, Handshake. We also found that a higher percentage of teleworkers post content and make comments. It is possible that both of these differences are attributable to teleworkers being remote and missing the rich communication of face-to-face interactions. During interviews, teleworkers stated that they found themselves using other communication mechanisms (such as instant messaging and e-mail) more than they had before; it therefore stands to reason that they may be more likely to adopt new communications tools early. This is supported by earlier findings showing that teleworkers are earlier adopters of technology that supports their work [15, 16] and that the willingness to become a teleworker is associated with a higher need for “innovativeness” [14].

In addition, we discovered that teleworkers tend to work more diverse hours than their non-teleworking colleagues. This was shown by our quantitative analysis, and a number of individuals spoke about this in the interviews as well. Some found their work extending into time with their family members. As one teleworker told us,

“Often I get in trouble with my wife when she finds me still at my desk and she thinks I should not be working.”

On the other hand, some teleworkers reported that they were now more strict with their working hours than they had been when they worked in one of the office locations. A teleworker explained his new working schedule since he transitioned to remote work:

“I am pretty strict about the hours I work… When I worked in the Bedford headquarters, I used to stay at the office until I finished what I was working on… I tried to keep work at work and not bring it home. Why the schedule difference [once I became a teleworker]? I… stop at a hard set time to see my grandmother, have dinner… [this routine] help[ed] me separate work and life because it would be too easy to work until 10 pm [when working from home].”

Others found that they had to shift the time they began their work earlier in the day in order to accommodate those in other time zones. One teleworker explained how being on the west coast of the US affects his interaction with his east coast colleagues:

“There is also the time difference. I have to get up at 4:30 am to have coffee to be awake for 5 o’clock meetings that someone schedules for 8 am on the east coast. It happens probably 10-15 times per quarter. I consider this part of the cost of being able to live where I want to live and still be able to support MITRE and our sponsors on the east coast.”

Still others felt that they had more flexibility over the time they spent and could take time off in the middle of the day for a doctor’s appointment or an errand. Regardless, we found that there was more variance in the times they started and ended their work.

Both our research, as well as the comprehensive review of best practices done by Koehne et al. [13], identified gaps in communication that are present in the teleworker environment and that some teleworkers do not feel as connected to co-workers as those in office locations. Teleworkers do not have the opportunity for “water cooler,” or hallway, conversations and have to find other ways to engage with their co-workers. As two teleworkers put it:

“It’s both an advantage and disadvantage. People don’t just drop by and interrupt, but you miss that social engagement.”

“You have to work harder to build relationships. You just want to jump in your car, shake their hands, and have lunch with them, but you just can’t do that.”

They also feel excluded from scheduled social events such as team lunches or office barbeques: one interviewee reminisced about door decorating contests as a seemingly insignificant activity that nevertheless brings people together. In addition, we found that teleworkers have to make an extra effort to keep in touch with people.

“It’s harder to stay connected to people. Out of sight, out of mind. You have to take that extra effort to initiate chat or call someone. You don’t run into people in the hallway or at lunch.”

Because of lack of face-to-face communication and the fear of “out of sight, out of mind,” teleworkers compensate by using online chat or picking up the phone more frequently than they might otherwise have had they been in an office location [9]. One person schedules what she calls “virtual teas,” which are informal phone sessions where she meets with colleagues to “catch up” and socialize regardless of whether they are working together. She described the genesis of how it evolved out of a face-to-face gathering:

“There are a couple of ways I keep in touch with people. Virtual Tea – we meet once a month or every couple of months for a half hour or 45 min just to chat. It’s like running into someone at the cafeteria but it’s more intentional… At one of the Virtual Teas, there are 3 of us. Our first one was in person at the cafeteria, and we all agreed to do it again regularly.”

When some teleworkers visit the office on business trips, they network heavily by scheduling informal lunches or making office “sweeps.” In general, teleworkers reported that communication was much more difficult from remote locations and that they had to be very deliberate and proactive. As one interviewee put it:

“I need to meet people and get to know them. It’s very difficult to do teleworking. I make a point to do office sweeps in the morning whenever I am visiting the office. The sweeps I do may be with people I have never met before. I might stop someone in the hallway. If I am working with someone I don’t know, I will look at their picture on the MII [MITRE’s intranet phonebook directory]. I like to see who they are so that I will recognize them when I see them. I never had to do this before but I found myself not knowing people I was working with.”

Given this tendency for teleworkers to feel more disconnected from co-workers, we hypothesize that individuals who work remotely on a regular basis would be more likely to use an enterprise social networking platform as an added way to communicate with others and create an online presence. Indeed, we found that teleworkers adopted Handshake earlier than others and had a higher contribution rate (both in terms of volume and frequency of postings).

One intriguing finding was that teleworkers were in a lower percentage of social groups than they were in work-related groups in comparison to the other populations. This was counter-intuitive to us as we expected teleworkers, who have less opportunity to socialize, to use this medium to socialize more than other populations.

One hypothesis for the lower social group membership for teleworkers is that there may be more of a need for teleworkers to cultivate an outward perception of productivity. Previous research [13] found that teleworkers have to do more to make their work visible. Since others cannot observe them doing the work, it has to be explicitly called out and shared by the teleworker or their management. Since co-workers and management are not able to observe teleworkers conducting their work, the teleworkers may feel the need to project a persona of “productivity” and produce more indicators of work. Participating in social groups may work against this intended persona and may therefore be avoided to a higher degree by teleworkers. Alternatively, it is less about creating a persona and more about spending time creating observable work deliverables rather than social conversation. We did ask some of our interview participants whether they did anything specific to cultivate their professional image and whether they felt that their co-workers’ perceptions of them had changed since they became teleworkers, but they did not specifically indicate that their lack of involvement in social groups had anything to do with maintaining their professional reputation at work.

Another potential explanation is that many social groups on Handshake are tightly connected to gatherings or location-specific activities, which require physical presence (e.g., softball or soccer teams, cycling enthusiasts, Frisbee players). Teleworkers, by virtue of the fact that they are not working from the same location as their co-workers, would naturally be left out of these organized activities. For example, while the cycling group on Handshake does contain general discussions about the safety of sharing the road or how to best ship a bike, there are frequently posts about planned group rides and local races in which a remote teleworker would not be able to participate. While this is a possible explanation, we would want to do further exploration to substantiate or refute it and understand why there is also lack of teleworker interest in other social groups that are not related to geographic location (e.g., gardening, vegetarianism, pet care, photography).

This study was a case study on a single company; therefore, the extensibility of the findings may be limited. In addition, The MITRE Corporation is unique in that it is not-for-profit and government-oriented, and most work is knowledge work – which lends itself nicely to telework.

7 Future Directions

This study provided insight into the working lives of MITRE’s teleworkers, and we plan on compiling relevant information into guidelines for teleworkers. With the aim of helping teleworkers maintain their productivity while retaining job satisfaction, we will focus on:

  • Strategies for managing schedule and availability,

  • Tips for keeping work life in balance,

  • “Social” conventions for handling customer communication, work assignments, and meetings with remote participants.

We also plan on creating guidance and best practices for managers and project leaders who deal with teleworkers on a regular basis as to how to better foster a working environment that enables teleworkers to be productive and happy. In addition, we hope to continue partnering with MITRE’s teleworkers to study challenges that ICT users face. Because of their reliance on collaboration tools, teleworkers have proven themselves to be valuable partners in our technology research program.

There are a number of areas that warrant further research. First, we observed a number of different types of teleworkers. While some literature [9, 17] does differentiate by frequency of telework (Full-Time, Regular, Occasional, etc.) and work location (home, office, field, etc.), we have not seen any differentiation in the literature for other dimensions of telework. For example, we interviewed some teleworkers who had spent many years in one of MITRE’s headquarters and, only towards the end of their career, did they become teleworkers. This was in contrast to others who were hired directly as teleworkers. The former had developed extensive networks from interacting with co-workers face-to-face, but the latter did not have that initial face-to-face benefit. Another difference was that some teleworkers had other teleworking co-workers nearby, while others were completely isolated. It is clear to us that there are multiple, varying dimensions of telework. We believe that additional development of the types of teleworkers and their features may merit further exploration – particularly if we can discover differences in productivity or work satisfaction.

Work/life balance for teleworkers is another area that we would like to explore further. While we did see teleworkers working an extended range of hours, we do not know if this resulted in working more hours in total, or whether this was simply the result of their days being punctuated by other responsibilities (e.g., child care).

Lastly, our finding of social groups having lower teleworker membership raises a number of questions. We proposed two potential explanations for this: one that pondered whether social groups are more geographically based and are therefore less relevant for teleworkers, and a second that teleworkers are concerned about creating a persona that shows them as “business-like” rather than “social” and they therefore refrain from interacting in more social venues. Further research would be required to determine whether either of these theories explains this phenomenon – or to identify an altogether different explanation.