PART 1: EXECUTIVE SUMMARY
The purpose of this paper is to articulate the manager and employee work practices that are
most important for the success of distributed work groups of various kinds. Its focus is the
interactive work of groups comprised of managers and employees who are in some manner
distant from each other. This paper aims to provide answers to the question "What do we do?"
to make distributed work situations work well.
The paper is based on a review of a wide range of academic and practitioner literature. It is
also based on the notion that desired group outcomes depend on certain “Success Factors”
being in place. It is these Success Factors that the Work Practices achieve. Nine are identified,
and when in place, help create employee engagement, productivity, satisfaction, and
achievement of business objectives. The literature strongly indicates that these Success
Factors interact with each other, positively reinforcing each other’s presence and negatively
reinforcing each other’s absence.
With Success Factors identified, Work Practices gleaned from the literature are aggregated and
divided into Manager and Employee practices. Eight categories of Work Practices emerge. The
work activities within these categories support achievement of at least one but more often
multiple Success Factors, and answer the question “What do we do” to make distributed work
The paper also identifies a variety of distributed group variables that influence what Work
Practices are needed in different situations: i) position within a group, Manager or Employee; ii)
type of work group, Operational Group or Project Team; and iii) type of group distribution, Local
The paper comes to several main conclusions:
The need for the Success Factors and Work Practices identified is driven primarily by the rising dominance of knowledge work. Work force distribution is one significant characteristic of knowledge work; knowledge resides with people worldwide. The practices identified are applicable to all knowledge-based groups, distributed and collocated.
With distance, the importance of certain of the practices increases and the risks of poor implementations grow. Most important are:
Formally engage in the creation of group expectations and agreements that will govern group members’ work together. The practice of creating a formal structure sets the stage for all the planned, unplanned, and spontaneous interactions critical to group work.
Intentionally create opportunities for group members to learn about each others’
experiences, work situations, and personal information that will affect the group’s ability
to interact effectively.
Engage in technology utilization practices that facilitate frequent, broad communications.
Utilize formal performance management practices that “close the loop” on all the
expectations and agreements created at group formation and modified later as new
information and circumstances arise.
Locally Distributed groups have advantages that should make adoption of this mode of work
widespread, to gain the benefits of enabling productive work from multiple locations while
retaining the benefits of physical proximity and avoiding a variety of organizational and
employee costs associated with traditionally collocated groups.
PART 2: INTRODUCTION
The purpose of this paper is to articulate the management and employee work practices that are
most important for the success of distributed work groups of various kinds.
In recent years as technology, markets, and other forces have allowed or required organizations
and teams to become more distributed, the benefits and pitfalls of distributed work groups have
become better understood through a body of research and practice. Some of the research
focuses primarily on the difficulties of distributed work. Other research, by both academics and
practitioners, sets forth a number of recommendations to overcome the challenges and
recognize the advantages of distribution. Taken as a whole, this body of research provides a
fairly comprehensive guide for distributed work groups.
Several aspects of the published research, however, are not ideal for those looking for a guide
to “What do we do” as managers or members of distributed work groups:
Most of the academic research looks at one attribute of distributed teams at a time (such
as “trust” or “conflict”), so that a manager needs to review a wide range of literature to
gather all the important guidance needed.
Research recommendations often mix “enabling conditions,” “practices,” and “supporting
infrastructure” in a way that makes it difficult to tease out answers to “what do we do.”
Much of the research (particularly academic research) seems to assume one type of
distribution of one type of group: a project team distributed across a geography that
makes collocated presence rare. While a Globally Distributed Project Team probably
represents the most difficult circumstance for distributed work, and is worthy of the
research that has been done, in reality many types of distributed groups exist, each
varying in its purpose, scope of geographic distribution, ability to have physical copresence,
and permanence of membership. All of these variables affect distributed
group work practices.
Much of the research done and tools (such as training tools) developed by practitioners
are either proprietary or in a form that is not readily accessible by others.
This paper attempts to aggregate the research from various sources and organize the findings
in a way that recognizes the diversity of group types and the differing roles of people within the
groups. The paper also suggests clear “buckets” for the different kinds of research findings and
recommendations, so that enabling conditions (such as “trust”) and work practices (such as
what needs to be done to create an environment of trust) are more clearly articulated and
hopefully more easily grasped.
A secondary purpose for this paper is to set the stage for the development of guiding
specifications for the design of systems of place, technology, and policy infrastructure
specifically aimed at supporting distributed group work practices. While specifications and
designs are beyond the scope of this paper, we will suggest how an understanding of distributed
group work practices is the essential basis for specifications and design.
Resources and References
This paper is not the product of original research. It is an essay that draws on a wide range of
academic research, the research and activities of practitioners in the course of developing new
types of work environments, and on summaries of research and literature compiled by
NewWOW. This paper does not claim to draw exhaustively on all available literature and
resources, but hopefully draws on resources sufficiently to propose in concrete terms the work
practices needed for successful distributed group work of various kinds.
If statements are made that are hypotheses, or are opinions based on work experience, I have
tried to clearly note the statements as such. A full list of references used is included at the end
of this paper.
My background and experience supplement the information gleaned from the literature reviewed
and creates a perspective through which the information is interpreted and presented. From
1988 to 2006 I worked at Sun Microsystems, Inc., a network computing company, creating work
environments for Sun’s employees around the world. In the first years at Sun I launched early
research efforts to understand how to support groups engaged in a wide variety of knowledge
work. I co-founded and was Vice President of Sun’s iWork (now Open Work) group, which
designs integrated systems of physical places, technology tools, and organizational policies to
support Sun’s increasingly mobile and distributed employees. Over the years we worked with
(and I continue to work with) a number of partners to explore how to support such a work force,
including Joe Ouye and the Workplace Productivity Consortium, Stanford University’s Center for
the Study of Work, Technology, and Organization, the Institute for Research on Learning, our
key consultng partner Iometrics, and of course NewWOW. My experience at Sun and work
with partners provides a lens through which I have reviewed the literature. My hope is that this
lens has been clear enough to allow objectivity in the review.
Focus of Paper
Our focus is the interactive work of groups comprised of managers and employees who are in
some manner distant from each other. This paper aims to provide answers to the question
"What do we do?" to make distributed work situations work well, a question asked by both
managers and individual participants in distributed work groups. What members of a distributed
group need to do is influenced by several variables:
Type of Distributed Group
Operational Group: A group that is defined by reporting relationships, often referred to as a
“department” within a larger organization. Operational Groups are generally created based
on geographic organization (such as “European Sales”), discipline (such as “Marketing”), or
customer focus (such as “Financial Services” or “Government”). Operational Groups are
often characterized by relatively long duration of membership, with members not necessarily
working together day to day on a common effort, but expected to share work processes,
knowledge, and attainment of overall department goals.
Project Team: A group that is defined by assignment to a specific design, development, or
implementation effort. While some Project Teams can be a subset of an Operational Group,
more often membership is drawn from multiple Operational Groups within a larger
organization, and often includes members from outside the larger organization. Duration of
Project Team membership is generally limited to the duration of the effort that is the Project
Scope of Distribution
Local Distribution: Members of a group are geographically proximate, such as in the same metropolitan area, but work from locations other than the group’s main location some or much of the time. In this paper, Local Distribution means that physical distance is not an
unreasonable barrier to physical co-presence, and as a result members can be collocated at least some part of each week.
Global Distribution: Members of a group are geographically distant, distributed across geographic areas and time zones that make travel to a common location time consuming and costly. As a result, physical co-presence is limited to specially planned events several
times per year.
Often, a group will have membership that represents both of these types of distribution.
Position Within The Group
Within an Operational Group:
Employee who reports to the group manager (or a lower level manager)
Within a Project Team
Employee who is a Team Member, but probably does not report to the Project Manager
This paper takes into account type of group, type of distribution, and type of position in
identifying distributed group work practices. But first, we need to understand what these group
work practices aim to accomplish, which are the “enabling conditions” for success. We call
these enabling conditions “Success Factors,” the factors that can lead to group productivity,
employee satisfaction and engagement1, and achieving the work output that is the reason for
the group’s existence. The relationship between Work Practices, Success Factors, and higher
order outcomes is shown in Figure 1, and serves as the organizing framework for this paper.
The diagram shows a left to right flow of impacts: Work Practices impact Success Factors that
impact higher order desired outcomes. We can also guess that when desired outcomes are
achieved, Success Factors will be positively reinforced -- and if outcomes are not achieved, they
will be diminished -- diagrammatically shown with the dashed line.
Based on the breadth of literature reviewed, and without minimizing the issues associated with
distance and group distribution, it appears that the major impetus for development of most of the
practices shown in this paper is not the increasing distribution of work groups, enabled by
technology and necessitated by a range of market and other factors; rather the main impetus
has been the emergence over the past fifty years of knowledge work as the dominant type of
work in many countries around the globe. Peter Drucker coined the term “knowledge work” in
his book The Landmarks of Tomorrow (1959). In The Effective Executive (1967) he made the
claim that knowledge workers with specialized knowledge cannot be managed, only encouraged
and enabled to do their best work. Encouraging and enabling groups to create and distribute
new knowledge in the form of new products, services, and processes are at the heart of the
Work Practices described in this paper.
Earlier than Drucker, in 1945, the economist Frederick A. Hayek wrote a “seminal article, ‘The
Use of Knowledge in Society’.” (Jensen and Meckling, in Knowledge Management and
Organizational Design). In this article, Hayek wrote of the “problem of the utilization of
knowledge that is not given to anyone in its totality.” Hayek’s insight, according to Jensen and
Meckling, was that an organization’s performance depends on placing decision-making authority
with the person or people whose knowledge is important to those decisions. Hayek argued that
the distribution of knowledge in society calls for decentralization.
Decentralization, mobility, and specialization of knowledge are the primary drivers of the work
practices that are so different from the practices of the centrally planned and controlled efforts of
the industrial era. With decentralization, mobility, and specialization of knowledge come groups
and the work practices associated with groups; and with groups come the needs to establish
common goals, common expectations, agreements on how to do things, who will do what, how
to share knowledge and other resources, how to manage conflict, how to adapt to different
circumstances, how to evaluate performance; in short, a new set of work practices. To
understand in more depth what work practices are needed, we now turn our attention to the
Success Factors that the practices need to create and maintain.
PART 3: SUCCESS FACTORS
As noted above, a distributed group’s Work Practices are those actions that result in creating
and maintaining a set of Success Factors that are necessary to achieving higher order
outcomes such as engagement, productivity, satisfaction, and the specific outcomes that are the
reason for the group’s existence. Uncertainty in how to achieve and maintain the states
described by these Success Factors is the frequent concern of those in distributed work group
situations, and is often the reason managers (particularly mid-level, front line managers) resist
adoption of distributed work situations even when doing so would otherwise help their groups be
Interestingly, the Success Factors identified in the literature about distributed work also appear
in literature about any team work that involves knowledge, ideas, and information: “Intellectual
Teamwork” is the phrase used by Galegher et al in their 1990 book that explores the
“Foundations of Cooperative Work.” So, these Success Factors are not unique to distributed
teams. But whereas many managers seem confident that they know how to achieve these
conditions in a traditional work situation, some of that confidence is lost when distance is
List and Description of Success Factors
The Success Factors that need to be created and maintained through the varying activities of
Managers and Employees, on behalf of their groups, are:
An Environment of Trust
Managers: “I trust my employees”
Employees: “I am trusted by my manager and my colleagues”
Managers and Employees: “We are able to express and discuss disagreement and
alternative points of view and resolve any associated conflict”
Employees: “I know I can rely on my team members for help and expertise.”
Control of Outcomes
Managers and Employees: “I know that the outcomes for which I am responsible and
accountable are in control”
Employees: “I know the areas for which I have decision authority”
Clarity of Expectations
Managers: “I am clear on this group’s charter and goals, and how they fit with the larger
Employees: “I know what my group’s charter is, what I am accountable for, what is
expected of me, and whether my manager and others believe I am meeting those
Employees: “I understand the roles and responsibilities of each team (or group)
member; I understand why each person is on the team/in the group”
An Environment of Support
Employees: “I am encouraged to do my best work, and recognized for my
Employees: “We help each other through difficult times”
Employees: “I have colleagues with whom I can share successes and challenges”
Ready Access to Material Resources
Employees: “The material resources (spaces, technologies, tools, travel) we need are
funded and available for our use”
Employees: “We know how to use the resources available to us”
Group Identity, Cohesion and Commitment
Managers and Employees: “I belong to the group and I am committed to its success”
Managers and Employees: “We share a high commitment to our effort and share a set of
Employees: “I am well informed, up to date, and included in group discussions”
Employees: “I am respected by my colleagues: my voice is heard, and my opinions
An Accepted Set of Group Processes and Protocols
Managers and Employees: “We understand how decisions are made and how decision
authority is delegated”
Employees: “We agree on how to ”
An Environment of New and Shared Ideas and Knowledge
Employees: “We have a free flow of ideas and conversation, we share knowledge easily
and widely, and discussions are broadly participative”
Employees: “We are open to hearing and discussing new and diverse ideas, and
challenging ‘accepted’ knowledge”
Employees: “Information and ideas are the domain of all, not a select few”
Efficient Use of Time
Managers and Employees: “Engaging with group members (for a wide range of
purposes) is easy and seamless”
Managers and Employees: “I know who I can turn to for particular knowledge and
Managers and Employees: “We are able to efficiently use the technologies that allow us
to engage over distance”
Discussion of Success Factors
How each of these Success Factors enables desired higher order outcomes is discussed in the
sources cited at the end of this paper. A thorough review of why each is important is not
repeated here. How the Success Factors (SFs) may interact is, however, worth brief
Based on review of literature, as well as the author’s experience, it appears that most of the SFs
interact with one or more of the other SFs in a positive way: the more (or less) one is in place,
the more (or less) the other will be in place. When in place, the interacting SFs act like a
reinforcing virtuous cycle. As a system, they represent a certain mindset or “culture” about the
nature of work. A command and control culture, for example, would probably not value this set
of SF’s. The highly interactive nature of SF’s may help explain the difficulty of changing an
existing system of work. If, for example, a group is newly distributed (as might happen with an
acquisition of a small, distant business, or if one or more of a group moves to a new geographic
location), the potential diminishment of one SF, such as Group Identity and Cohesion, will likely
diminish the group’s overall Environments of Trust, Sharing of Ideas and Knowledge, and sense
of Employee Support; and the diminishment of trust will further impact other SF’s. The
introduction of a significant change, such as work force distribution, into an existing system of
work requires attention to the whole system of SF’s and supporting work practices. Such an
effort is difficult for many organizations to absorb, if their system of work is not already well
established to support knowledge work in general.
Figure 2 is an estimate of how a SF interdependency “Map” might look. Although the literature
reviewed is not explicit about how all the SFs interact, the literature that discusses Trust and
Expectations clearly indicates that these SFs have unique positions in our map. Clarity of
Expectations is the important catalyst for achieving all other SFs. An Environment of Trust is the
result of a number of the SF’s being in place, and in turn Trust reinforces those and other SFs.
The Expectations and Trust SF’s along with three others may form a “core” set of SF’s. The
additional three are: Accepted Processes and Protocols, Control of Outcomes, and An
Environment of New and Shared Ideas and Knowledge. It may be that the Work Practices
needed to achieve these core SF’s are the most critical for a distributed group to adopt.
Figure 2 illustrates the centrality of Clarity of Expectations and an Environment of Trust in the
Success Factors map as “starting” and “convergence” points, and how the set of five SF’s
create a “core.”
Figure 2: Success Factors Interdependency Map
Large arrows indicate centrality of the Expectations and Trust Success Factors.
Small arrows indicate other important interdependencies.
An interesting exercise is to change one of the nodes in this map to "no" or "low". For example, change the presence of An Environment of New and Shared Ideas and Knowledge to lack of that presence, and picture the ripples through the system of SF's.
Clarity of Expectations
This diagram implies that the "starting point" for the network of SF's is Clarity of Expectations. Clarity of why a group exists and what each person is expected to contribute starts the system of SF's rolling, towards both an Environment of Trust and Efficient Use of Time.
John Gabarro of Harvard University in his review of literature on working relationships (in Intellectual Teamwork) suggests that "the structuring of expectations is the single pattern that contributes positively to productivity and satisfaction." These expectations address performance, goals, and each party’s role in a working relationship. He further states that
mutual expectations “are more typically worked out over time during a succession of routine
interactions....” However, in today’s relatively fast paced world, particularly for project teams
with limited duration, the time may not exist for gradual emergence of expectations; and
distribution of group members is likely to be less forgiving of early misalignments in
expectations. Rather, Clarity of Expectations is partially dependent on formally developing an
accepted, explicit set of protocols and processes through concrete, early group discussion and
negotiation. Once sufficient trust is in the system, group members can allow those protocols
and processes to evolve and adapt to new circumstances, allowing modifications to
expectations to emerge, as Gabarro suggests.
Clarity of Expectations in itself does not create healthy distributed group work, but without clarity
group members will likely be in a state of ambiguity at best, and in conflict with each other over
roles and responsibilities. Commitment to the activities needed to build the other SF’s will be
difficult without clarity about why the team is important, what it needs to achieve, and how
people will work together.
Environment of Trust and Control of Outcomes
Trust is both an outcome of having a number of other SFs in place and a contributor to those
SFs as well. Because of the centrality of Trust in the system, and because reduced trust is one
of the most frequently mentioned concerns by practitioners as well as a common finding by
those doing research on distributed teams, some further discussion about Trust is useful. In
addition, Control has some complexity worth noting.
Several researchers identify at least two different kinds of trust: a personal relationship type of
trust, and a more work related type of trust. Personal relationship trust (“personal trust” or
“emotional trust”) refers to trust in personal characteristics such as being honest, open, and
supportive. Having personal trust means, for example, that group members will not be overly
concerned with the risk of disagreeing with each other on an issue; they will trust that they can
have an open discussion about the disagreement and reach resolution.
Work related trust (or “cognitive trust”) refers to trusting that a person will deliver on
commitments made. Having this type of trust means that once commitments are made to each
other, group members perceive little risk to those commitments.
Drawing on research by Griffith, Mannix, and Neale (in Virtual Teams that Work), it appears that
work related trust is closely associated with minimizing the risks associated with what Griffith et
al call task and process conflict; and personal trust is associated with minimizing the risks
associated with relationship conflict. Their research data suggest that for distributed teams, the
frequency of relationship and task conflict is about the same as for traditional collocated teams.
However, process conflict can occur much more frequently in distributed teams. This indicates
that distributed teams need to pay particular attention to the risk of process related conflict, and
need to build commitments to process and protocol agreements which in turn will help build
work related trust3.
Work related trust may be more important than personal trust to Project
life span, whereas Operational Groups’ achievement of multiple and
longer life span may be as much or more dependent on personal trust.
Control of Outcomes is also strongly correlated with Trust. Successful Control of Outcomes in
knowledge organizations cannot mean what Bradford and Cohen refer to as “Heroic”
Management, in which the manager is the center of all activity; nor even what they call
“Manager as Conductor,” in which the manager is the source of all coordination. Rather, Control
of Outcomes as a Success Factor means that a manager has successfully developed
subordinates’ capabilities to share management of the group’s performance, using what
Bradford and Cohen call “Manager as Developer” practices.
The more an Environment of Trust is in place, the more a shared management environment will
be evident with more decision authority and control of outcomes delegated to group participants.
A Sun study found that this in turn will provide managers with confidence that the outcomes for
which they are accountable are indeed under control. In Sun’s terms, managers will feel more
comfortable in adopting “Management by Believing” that good work is being done, rather than
“Managing by Seeing” their employees in a work location, and thereby assuming that only then
is work being accomplished. (Sun Manager Focus Groups, 2003).
The Sun study indicates that in general managers’ concerns are focused on group outcomes
more than on individual outcomes. Thus we can hypothesize that work-related trust -- delivering
on commitments -- is what is primarily needed to build a manager’s trust in employees. When
this trust is in place, managers can focus on what they are uniquely positioned to do for the
group, “working the boundary conditions” to gain needed resources and cooperation from senior
managers and other organizations, rather than focusing on the day to day activities and
coordination of the group.
However, from an employee’s perspective personal trust in one’s manager may also be
particularly important to enabling constructive discussions about ambiguity, conflict, new ideas,
and personal development. Thus, an employee may value both personal and work related trust,
since personal trust with one’s manager allows important discussions to occur, and work related
trust allows delegation of authority that will enable both an environment of employee support
and growth, and, through delegation of decision-making, better use of time.
The SF Interdependency map shows a number of routes to trust. One might think that this
represents some redundancy: if one route is deficient, another route can be used. But more
likely, the multiple routes to trust represent multiple vulnerabilities: many ways to erode trust.
The literature indicates the importance of all the enabling conditions that interact with trust, so
the practices that allow creation and maintenance of those conditions deserve close attention.
PART 4: WORK PRACTICES
With the Success Factors listed and described, we can turn our attention to the Management
and Employee Work Practices needed to achieve these Success Factors. This section of the
paper describes what these Practices, or work activities, are, which are answers to "What do
we do" when in a distributed work situation.
Work Practices Overview
The literature reviewed describes a wide range of work practices needed to achieve the SF’s.
This range of practices appears to fall into a handful of activity types:
Information and Idea Sharing
There is clearly some overlap between these types of activities (for example between
Communications and Sharing), but hopefully organizing the practices gleaned from the literature
in this way provides acceptable simplification for presenting the material. Tables in Appendix
Parts 7a, 7b and 7c provide complete listings of the Work Practices that comprise these activity
types4. Here, we provide an overview of each activity type.
Group Formation practices are among the most critical for a distributed group?s ability to achieve
multiple Success Factors. Literature about collaborative group work discusses the great
importance of establishing stable group sponsorship and leadership; compelling, challenging
missions and sets of goals for teams, and a sense of urgency in accomplishing those goals;
understanding what a group?s “task boundaries” are; clear understanding of who comprises the
group, and what the valuable roles are for each member; members? understanding of each
others? work situation or “context,” including personal circumstances and cultural differences
that may have important impacts on work behavior; understanding what knowledge, experience,
and network each member brings; agreeing on the processes, protocols and tools that the
group will use to meet, communicate, share, and make decisions; and creating a common
understanding of performance standards (which will likely evolve as tasks progress).
For collocated groups, particularly Operational Groups with membership stability, some of these
understandings can emerge over time. For distributed groups, in particular Project Teams with
limited time frames, Group Formation activities establish, or at minimum kick-start, the
conditions needed for establishing “swift trust” (Mannix, Griffith and Neale in Distributed Work)
between team members. Group Formation activities for distributed groups, including annual
group “renewal” for Operational Groups, involve some “slowing down” early in order to “speedup” later in the groups? work (Griffith et al in Virtual Teams that Work). Most important, Group Formation is a set of explicit, planned, intentional activities aimed at accomplishing early the mutual expectations and agreements that many collocated groups allow to emerge, or assume will emerge, implicitly and “organically” over time. And, it is very important that the expectations and agreements that have to do with how work will be done -- processes, protocols, and so on -- be determined by the group members, not the Manager. Effective teams of all kinds, according to Mannix et al (in Distributed Work) are those for which the Managers set “clear and engaging direction” and the team members determine the means to achieve their goals.
Communications practices involve communications between Manager and Employees; between
Employees within the group; and between Manager and the larger organization. These
practices include both formal (planned) and informal (unplanned) activities.
Of all the communications practices discussed in the literature and listed in the Tables in the
Appendix, the practices that stand out the most are those that create “communication
density” (Kanter, in Knowledge Management and Organizational Design). Using the
accessibility and availability protocols and technology agreements reached during Group
Formation, it is important that group members communicate frequently and broadly with each
other, about matters large and small. Some of this communication will be planned: in regularly
scheduled meetings, for example, using good distributed meeting protocols5. Other
communication needs to be unplanned: quickly accessing each other with questions and
coordination issues as well as just checking in with each
other, letting others know what’s “going on,” share gossip,
discuss successes and failures. Similarly, it is very
important for Managers to check in with the group, both as
a group (such as during regular meetings) as well as with
individuals to see how they are doing, assist with a
problem, and so on. For groups chartered with the creation
of new products, services, or processes, frequent informal
communications, both synchronous and asynchronous, are
critical to enabling the “surprise and unexpected” that
“lubricate innovation and creativity” (discussion with Joe
Ouye). Distributed groups need to become comfortable using technology to mediate these informal communications6.