An Architecture and Business Process Framework for Global Team Collaboration

Tools for remote team collaboration within businesses have been available since the mid-1980s. Two opposing trends cause complete collaboration solutions to remain elusive. On the one hand, core tool capabilities are developed as point solutions, and then extra functions are added. These added functions may not integrate well with or be as fully developed as the core functionality. On the other hand, enterprises are rapidly globalizing and becoming more dependent on comprehensive collaboration applications to coordinate distributed teams. This means that overall productivity is affected by how well tools, processes, and capabilities are integrated; the tools should not be just a collection of separate features/functions. An audit of collaboration tools used at Intel showed both overlaps and gaps between remote tools and day-to-day activities of workers. When an employee has so many tools to choose from and furthermore, works on multiple teams, the choices become overwhelming and confusing. The underlying architecture of a realistic solution to these overlaps and gaps must provide integration interfaces within the team collaboration environment, and to other business applications, information technology services, and infrastructure. In this paper, we describe a multi-level approach to integration, and we discuss unique findings about Intel’s remote teams that justify our model. An essential element of progress towards the goal of an integrated solution will be the deployment of enabling platforms and the likelihood that these practical, indeed necessary, innovations in collaboration will also provide market pull for Intel’s core products. By identifying and addressing our own needs, we can also provide solutions for a significant percentage of the Fortune 500 market that engages global workforces for knowledge work. INTRODUCTION Global expansion, outsourcing, competitive pressure to do complex projects more efficiently, and increased focus on work-life balance drive the need for employees to collaborate more effectively. Globally expanding companies face challenges in melding multiple cultures with diverse values, histories, and perspectives. This can make it hard for individual employees to understand their colleagues even when they share a corporate culture–especially when a company is functionally distributed. Today, we use information systems to improve the productivity of individuals or to automate tasks. However, these mainstream information systems do little to improve the ability of groups of people to work together on collective tasks such as collaborative problem analysis, idea synthesis, decision making, design, conflict resolution, and planning. Team productivity and performance has the potential to yield exponential results due to synergistic factors, knowledge creation and construction, diverse perspectives, and coping with complexity. Almost every business process or project involves some form of collaboration and coordination between participants. Globally dispersed teams incorporate talents from different locations, and key team members can be chosen for their proximity to important customers and other stakeholders [1]. With the right collaboration tools, companies can become more agile and reduce product time to market [2]. Teams that effectively collaborate avoid or significantly reduce the following cost factors: • Time to market: Cost of not meeting market window, loss of competitive advantage. • Time to information: Project delays due to lack of information or incorrect information. Intel Technology Journal, Volume 8, Issue 4, 2004 An Architecture and Business Process Framework for Global Team Collaboration 374 • Cost of duplicate projects: Unintended duplication of effort. • Cost of poor coordination: Increased risk of severe product flaws and recalls. • Travel and relocation: Remote coordination instead of face-to-face meetings and co-location. • Opportunity cost of intellectual capital: Knowledge worker hours can produce more variation in value than manual worker hours; teamwork hours can produce exponentially more value than individual worker hours. Remote collaboration software and related research have been under development since the mid-80s [3]. In most cases, the market has offered clusters of point solutions that represent their own core capabilities. Not only do these point solutions fail to encompass complete sets of collaboration needs, they also tend to embody outdated models of how corporations work that no longer apply in the global enterprise. Moreover, team collaboration products on the market do not support enterprise-scale multiteaming; lack interoperability of needed applications; do not support business process and application integration; nor do they promote fluid switching between asynchronous and synchronous collaboration modes within a single environment. End users must learn different tools for different collaborative activities and move information across multiple environments. At the scale of an 80,000 person enterprise with two-thirds of the employees engaged in multiple teams, these constant shifts in applications and environments are counter-productive. Most collaboration tools lack rich, expressive, userfriendly interaction models, the kind that are needed for groups that rely on distance collaboration full-time and that include members who may never have met face-toface. For this reason, they may fail to attract pervasive recurring usage. To address these challenges, Intel’s IT Virtual Collaboration Research Team [VCRT] decided to do the following: • Survey Intel’s “virtuality” on five dimensions: time, space, business unit, media, and culture. • Create a baseline of related external research and current collaboration tool use at Intel. • Identify the “desired user experience.” • Design user-oriented solution concepts for Intel and similar organizations. • Define a service-oriented architecture for team collaboration. • Specify enabling IT infrastructure dependencies. The baseline and survey work validated the first-hand observations of team members and led to a definition of the desired user experience. By starting with the desired user experience as a goal, the team avoided the trap of thinking in terms of predefined capabilities. Instead we worked from the experience to identify supporting capabilities. New capabilities offered in the collaboration environment map to the unmet needs identified in the virtuality survey and baseline work. For example, 64% of our employees belong to upwards of three or more teams; yet, multiteaming is not addressed by existing commercial tools. Other unique capabilities in our concept vision (see section entitled Visionary Concept) include individual and team workspaces for coordinating among multiple teams, asynchronous workspaces for tracking collaboration across time zones, and expressivity for social bonding when employees don’t meet face-to-face. The design also addresses ease of use and navigation via an objectoriented 3D graphic desktop. These new capabilities are as important as existing ones such as document storage and shared visual communication. However, there was a new focus on integrating all user needs into an interoperable collaboration environment. The VCRT has made significant progress towards understanding and measuring virtuality, and in developing the overall concept design. Because of the interdependencies among the user interface, business process, applications, architecture, and enabling infrastructure layers, the team continues to explore these layers and their interconnections. Future research will evolve the asynchronous and mobility capabilities, prototype an interactive environment, validate user acceptance, and increase understanding of enabling technologies, architecture requirements, and feasibility. INTEL VIRTUALITY DATA A 2003 survey conducted at Intel Corporation [4] with 1260 respondents created an index to measure Intel’s degree of team “virtuality.” The purpose was to identify the potential payback of a radically new collaboration solution. Virtuality was defined by an initial set of five “discontinuities”: time, space, organization, culture, and media [5]. Two new discontinuities added to the construct of virtuality were multiple teaming and differences in tools and practices. Three factors emerged in analysis: team distribution, variety of practices, and workplace mobility. Team distribution measures the degree to which people work with others distributed over different geographies and time zones. Variety of practices measures the degree to which employees experience cultural and Intel Technology Journal, Volume 8, Issue 4, 2004 An Architecture and Business Process Framework for Global Team Collaboration 375 work process diversity on their teams. Workplace mobility is the degree to which employees work in environments other than regular offices, including different Intel sites, home, travel routes, and places outside the company. On the one hand, we found that being geographically distributed in and of itself had no impact on team performance as measured by conformance to Intel values of discipline, quality, customer-orientation, risk-taking, results orientation, and great place to work. On the other hand, lack of shared work practices and structure, and workplace mobility negatively impacted performance. Cultural differences also posed a challenge to perceived performance. 71% of employees work on teams with people of a different culture or with people who speak different native languages or dialects of English. Thus, cultural differences are a key factor in any collaboration tool strategy. In interviews with employees about team processes a consistent complaint was that the variety of processes, particularly the variety of tools used in a process, was a main source of frustration. (Variety of tools means employees need to use different tools and processes for different projects, or they have to resort to different tools to accomplish a single task.) The data also pointed out that two-thirds of Intel employees belong to three to ten teams simultaneously. This situation compounds the problem of different tools and processes, as employees must often switch tools for the same activity when going from one team to another. For instance, documents may be stored in a collaboration portal, on a Web site, on a shared drive, sent as e-mail attachments, or simply shared on the desktop in real time. Multiple teams times multiple methods compounds the negative impact of diverse practices. The survey was conducted to answer questions that came up during the discussion of collaboration tool design to test whether our observations could be generalized across the full spectrum of Intel job types, ranks, business units, and geographies. We were surprised to find that virtuality experienced on the factory floor was similar to that of non-factory knowledge workers. The data provide strong support for our envisioned collaboration solution, which provides a platform for the integration of different processes and tools. In 2004, we repeated the virtuality index survey and our results revealed that the five key metrics and associated indicators used for the virtuality index all increased. Some of these increases were marginal, while others were statistically significant. Figure 1 shows indicators with significant per-employee trending from 2003-2004. 0 20 40 60 80 100 e-mail discussions cross-cultural collab screen-sharing tools across time zones X biz group collab diverse tracking tools shared repositories extend days for X-site instant messaging collab X-site Percentage 2003 2004 Figure 1: Virtuality trending: 2003-2004 These changes have differing impacts on productivity. Our earlier study found that distance and crossing organizational boundaries alone had no effect on perceived performance, whereas using diverse tracking tools had a negative effect on performance. The latter has increased by 6%, indicating that the need for a common collaboration platform with shared business process tools is increasing. The growth in various available tools also denotes more cross-site work and uptake of existing technologies. Instant messaging in particular had just been introduced at the time of the last study and is showing nonlinear growth with respect to other indicators. This reinforces the VCRT’s identification of presence awareness and sociability tools as a key need of the global workforce. Overall metrics of team performance to Intel values (as perceived by the individual) have either stayed steady or improved (e.g., “work fairly distributed” has increased 10%, from 44-54%). However, one disconcerting finding is that project timeliness has declined from a 52% rating to a 48% rating. There may be many explanations for this change, but a likely one is that coordination is suffering from the increases in the other virtuality factors. If so, this is a problem that needs an immediate solution. These data confirm our initial hypothesis in tracking virtuality: it is increasing rather than static. An additional hypothesis is that there can be a critical mass threshold that would create nonlinear changes in the organization’s performance. Supposing this to be the case, then the increasing dependence on remote teams makes the need Intel Technology Journal, Volume 8, Issue 4, 2004 An Architecture and Business Process Framework for Global Team Collaboration 376 for effective collaboration support a core infrastructure requirement. This hypothesis could also explain why integrated tool sets have not reached the market already: the conditions that impact a company’s bottom line have not been fully present up until now. There is no indication that virtuality will decrease. Valuable research from the academic community is also being incorporated into our ongoing plans. In particular, the work of Carmel and Espinosa [6] has informed the team’s sense of urgency and research about time zone differences and the growing need for asynchronous teaming tools to meet the demand for a better coordination capability. Majchrzak and Malhotra [7,8] describe a range of team needs for both cognitive and social integration, as well as task execution, tying these needs to how IT departments can respond. Hinds et al. [9] have compiled recent case studies on global collaboration, and the VCRT expects further engagement with these researchers. As well, the user interface needs of asynchronous team tools, and their ability to maintain user engagement across very large time separations will lead the research into new areas of engagement involving theories of “flow” [10] and experiments with multi-user video games that show what visualizations and activities keep users engaged and bring them back to the online encounter. Some of these findings can be applied to the problem of working environments where analogous activities may occur [11, 12].