Communications of the ACM
Home/Magazine Archive/June 2009 (Vol. 52, No. 6)/Deriving Mutual Benefits From Offshore Outsourcing/Full Text
Virtual extension
Deriving Mutual Benefits From Offshore Outsourcing
The concept of outsourcing has been addressed from different vantage points by various researchers. Many researchers have examined the impact of outsourcing on employment. Wiederhold et al14 looked at intellectual property and tax implications when a software company operates in multiple countries. This article takes an entirely different perspective. What are new computer-based techniques that can be employed to yield innovative solutions that can benefit both developed and developing environments? What is the long-term vision for conducting work in a global economy? Is there an end scenario for offshore outsourcing?
In order to make predictions about the long-term implications of the information revolution, let us look at the impact of the industrial revolution. When machines were first developed, they were scarce and costly; accordingly, the owners strived to utilize them on a round-the-clock basis, and employees were directed to work extended hours: 14 hours or even longer!
Governments stepped in to create laws restricting the maximum number of hours worked each day, and companies responded by gradually implementing a three-shift system. The concept of shifts suited many categories of tasks, but was inappropriate for other categories. Agricultural implements and military goods could be produced quicker and at lower costs. But the shift concept was incompatible with activities related to many types of fine arts. If three artists worked in succession for 8 hours each on the same painting, then the master painter may consider the product to be worthless.
Raman Roy, an outsourcing pioneer, observes: "Distances have become meaningless now; geography has become history." Today, we can transfer knowledge-based tasks to colleagues located in other continents instantaneously and at virtually zero costs, enabling them to continue the task on an incremental basis, and to create a "hybrid offshoring" environment incorporating both onshore and offshore activities.4
Long-term Vision for Conducting Work in Global Economy
Consider the following scenario. A female white-collar professional works from 9 am to 5 pm in the U.S. At 5 pm, she transfers the incomplete task to a male colleague in Australia who works from 9 am to 5 pm based on the timing in his timezone. At 5 pm in his country, he transfers the task, as updated, to a colleague in Poland who does further enhancements over the next eight-hour period, and then forwards the work to the first professional in the US. When the latter comes in for work, she feels as if a magic fairy has been working for her while she was asleep!
The above work scenario leverages both spatial and temporal separations by involving persons in different countries and leverages the differences in time between these countries to enable round-the-clock operation.3,6 This concept is utilized today by distributed 24-hour call centers where incoming calls are automatically directed to the center that is operational at the particular time. The success of decentralized centers can be attributed to: voice-over-IP technology; inexpensive and widely available computer and communications infrastructure; absence of governmental restrictions on trans-border services; and the employees who present the appearance of a seamless company.
A call center represents structured work. At the other end of the spectrum is ill-structured work such as the types of tasks performed by President Barack Obama and other heads of states; we do not envisage the use of the proposed paradigm for such tasks. In between these two extremes, there are many semi-structured tasks that will be increasingly addressed using the 24-hour Knowledge Factory (24-hour KF) paradigm (see Figure 1). Only some of these endeavors relate to the computer industry, such as the management of global communications networks and the design of VLSI chips. The vast majority of them relate to other industries, including finance, marketing, medicine, logistics, and law. Computer professionals will create the infrastructure to support distributed workforces for such industries, creating a surge in demand.
Overall, we will see a major change in the motivations for offshoring: from economic drivers in the short-term to the creation of strategic global partnerships and 24-hour KF in the long-term. Over time, the geographic location for the place of work will be determined by such considerations, not by the proximity to the place of the sponsor of the work. This may also help to reduce the incidence of prostate cancer in men and breast cancer in women.15
Research and Prototyping Efforts
Our research in this area commenced with a sponsored project aimed at developing new approaches that would enable quicker development of satellites. We found that these satellites were being created as pieces of art, one at a time, with little knowledge being transferred from experience on one satellite to another. We explored ways in which the information could be captured on an automated basis, then refined and distilled before being presented to the designer of the next satellite.13
Initially, we attempted to transcend issues related to geographic separations. Subsequently, we found that overcoming issues related to temporal separations could yield greater benefits. At one major U.S. company, programmers worked for a month, and then sent the code to a different group, within the U.S., for testing. Then, someone mooted the idea of getting the testing done in India. Now, the programmers send the code for testing at the end of each workday. The testing is performed on the opposite side of the earth, and the results become available to the programmer at the beginning of the next workday.
The effectiveness can be enhanced further by using three teams as shown in Figure 2. In this figure, each team works on a different aspect of the same problem. A more authentic example of a 24-hour KF involves the same task being handled on a successive basis by different teams around the world. For work to be transferred from one center to the next, one-third of the globe away, the challenge is to develop techniques that will allow such transfers to occur quickly and intelligently. Between the times that a team relinquishes work and resumes work, two others teams have worked for 8 hours each. So, at the time of resuming work, the team needs to know what exactly happened in the intervening 16 hours; and quickly too!
At MIT, we developed a prototype, KNOWFACT (from Knowledge Factory), for sharing semi-structured knowledge.13 In a 24-hour KF, traditional face-to-face interactions cannot be utilized for making crucial decisions. In order to mitigate this problem, KNOWFACT was designed to calculate the utility of any proposed system change, in response to different "what...if...?" questions from the stakeholders. KNOWFACT comprises of a Decision Rationale Module (DRM) and a Decision History Module (DHM). DRM uses Web-based inputs from individual stakeholders to define the characteristics of the end-product or end-service, with details being added on an incremental basis. DHM automatically captures historical information on ongoing activities. The combination of DRM and DHM provides an evolving knowledge repository that contains all the essential information, including that from the most recent shift of work. DRM ensures that the opinions of individuals who are not currently at work get equitable representation as compared to individuals who are currently working. When endeavors are repeated, individuals are provided with comprehensive knowledge of previous efforts. This facilitates knowledge to be shared between teams working at different locations around the world.
In software engineering, the overall endeavor is decomposed, on a recursive basis, into a comprehensive set of modules and classes, each of which can be ideally comprehended by a single individual. This individual can be deemed to be the "owner" of the particular module, through the process of coding, testing, and maintenance. In a 24-hour KF, the single ownership model constitutes a performance bottleneck because work during 16 of the 24 hours of the day occurs at centers that do not possess direct, synchronous communication with the owner of the concerned module or artifact. With the conventional principle of one owner for each artifact, necessary changes must wait until the owner's site turns on, causing time delays for related activities at the other two centers.
To mitigate this problem of asynchronous delay, compounded by issues of differences in language and culture, we developed the concept of composite persona (CP). A CP appears as a single individual to the outside world, but is actually comprised of three persons, with one person at each of the collaborating sites. As each site turns on and off, the role of the CP moves from one member of the CP to the next. Different CPs handle different modules, though a one-to-one mapping between modules and CPs is not required. Each CP acts as a single entity with frequent communications between members within each CP, and less frequent communications with other CPs. The concept of the CP provides benefits in terms of: smoother handoff between neighboring sites; increasing trust between members of these sites; accelerating the process of convergence on issues needing decisions; providing 24-hour access to other developers who may have urgent questions on the particular module; localizing lateral communication between owners of different modules; and increasing organizational resilience in terms of loss of knowledge when a person resigns abruptly.
These ideas are incorporated in Multimind,1 developed by us at the University of Arizona. Multimind is governed by three design principles: automate everything that is possible; store everything that is possible; and provide the human operator with the key information only, with the remaining being accessible on an as-needed basis. Multimind stores all e-mail messages that the individual sends or reads, all Web pages viewed, and all searches that were initiated. All activities are time stamped and logged into a comprehensive knowledge repository that evolves over the lifetime of the project. When the current worker encounters a change made by the predecessor, the development system queries the knowledge repository to reconstruct the sequence of knowledge operations that were performed earlier.
Unlike current software techniques that focus on horizontal decomposition of tasks, Multimind is designed to facilitate vertical decomposition of work. Consistent with the theme of the 24-hour KF that emphasizes vertical decomposition, collaboration occurs in a "high cohesion" gear that involves sustained knowledge transfer and communication between members of the collaborating sites. All events, as well as all atomic operations of all project objects, are recorded using LifeStreams,2 a chronologically ordered WORM object database. A worker can view the rationale behind the state change by querying the knowledge repository based on the timeframe when the state evolved. The results, shown in Figure 3, are categorized by event and prioritized by relevancy.
Related technologies include: User-Monitoring Environment for Activities (UMEA)7 and Time Machine Computing.11
Commercial Harbingers of 24-Hour KF
Examples from industry of harbingers of 24-hour KF come in several flavors: some involve globally distributed work, others involve two centers, and still others involve more than three centers. The examples here are illustrative, not exhaustive, in nature.
In Massachusetts, McDonalds is testing a new way to take orders. You drive in and speak in your order. This order is taken by a human operator located in another state and connected to the test site via electronic links. Conceptually, an order placed at night could be taken by an individual abroad, working during day in that country. Similarly, Siemens AG uses automated tools that can be monitored and handled from any geographical location; the traditional factory environment is transformed to a virtual "knowledge factory' where operation continues without factory workers having to work the night shift at any location.
The mobile industry firm, WDS-Global, used extreme programming concepts, Virtual Network Computing (VNC), and video conferencing to support a round-the-clock software development project, with sites in U.S., U.K., and Asia. Initially, the handoff process consisted of the transmission of a daily work summary; later, additional items were included. The project faced difficulties created by cultural differences and by long delays when downloading version control software from remote locations. One key lesson from this project is that participating teams should be of equal size; otherwise the largest team tends to dominate.
Two parallel software teams were studied at IBM.12 All the members of the first team were located in US. For the other team, one-half of its members were based in India, and the other half in US. The two teams reported to the same manager, developed similar products, and were closely monitored in terms of use of email and other electronic communications techniques, adherence to schedules, quality of intermediate and final deliverables, and allied parameters. Based on objective measures, the overall performance of the two teams, over a one-year period, was virtually identical. In fact, the distributed team performed better in terms of: superior documentation and retention of history; enhanced ability to share short term tasks; more structured definition of work tasks and distribution of work items; and higher quality of knowledge repository created for recording rationale for all decisions.
Office Tiger, based in multiple countries in North America, Europe, and Asia, provides banking and financial services on a consultancy basis. It uses these distributed resources to complete one-third of its assignments in less than three hours! A sophisticated information system, called T- Tracks, is used to collaborate between disparate locations.
End Scenario for Offshore Outsourcing
In the above cases, the challenge for the computer community is to develop the IS infrastructure that can be utilized by these industries to enhance product and service quality, to lower costs, and to reduce turnaround times. The challenge is also to develop new ways for storing and leveraging process rules and knowledge so that one does not need to reinvent the wheel every 23 years.8,9,10 Ideally, one would like to use the same IS infrastructure to provide 24-hour KF capabilities to diverse arenas; in reality, the IS infrastructure may need to be adapted to address the specific requirements of different endeavors.
In order to manage these distributed knowledge factories, professionals are needed to supervise persons from diverse cultures and backgrounds. In China, Russia, Japan, and many other countries, most individuals grow up in an environment that is relatively homogeneous in terms of religion, language, culture, and ethnicity. Persons growing up in U.S., on the other hand, are exposed to greater diversity on some parameters; accordingly, they enjoy a potential advantage in terms of being trained to serve as managers of global resources, located anywhere in the world. However, this ability is critically dependent on the educational system providing exposure to history, culture, and other aspects of societies from different parts of the world. Unfortunately, some of the educational systems have opted to become more restrictive in terms of coverage of such issues. Countries that recognize this problem and opportunity will witness greater success in training their citizens for work in foreign countries. The evolving laws at local, state, and national levels will also play a role in the transformation process.5 Over time, the notion of nations will hold less significance than it does today, with individuals being trained for the global economy, not for work in a particular city or region. Similarly, many categories of work will be conducted in a seamless manner at multiple locations, without regard to spatial or temporal considerations.
Conclusion
Offshore outsourcing can lead to faster development of new products and services through the use of the "24-Hour Knowledge Factory" concept. This paradigm involves the use of three or more teams of workers located at different places around the world. Each member of the team works the normal workday hours that pertain to his or her time zone, and then a fellow team member located in a different time zone continues the same task. This hybrid offshoring model can lead to strategic and economic advantages to all participants,6 and will be utilized by an increasing number of diverse industries over time. Computer professionals will enjoy new opportunities in designing new infrastructures to support the concept of distributed work in other professions.
References
1. Denny, N. et al. Hybrid offshoring: Composite personae and evolving collaboration technologies. Outsourcing and Offshoring of Professional Services. Arnar Gupta, Ed. IGI Global; Hershey, PA, 2008. 270286.
2. Fertig, S., Freeman, E., and Gelernter, D. Lifestreams: an alternative to the desktop metaphor. Proceedings of the SIGCHI conference on Human Factors in Computing Systems, 1996. 410411.
3. Gupta, A. Expanding the 24-hour workplace. The Wall Street Journal. Sept. 15, 2007.
4. Gupta, A. Outsourcing and Offshoring of Professional Services. IGI Global, Hershey, PA, 2008.
5. Gupta, A. and Sao, D. Anti-Offshoring Legislation and United States Federalism: The Constitutionality of Federal and State Measures against Global Outsourcing of Professional Services. The Texas International Law Journal 44, 3 (forthcoming).
6. Gupta, A., Seshasai, S., Mukherji, S., and Ganguly, A. Offshoring: The transition from economic drivers towards strategic global partnership and 24-hour knowledge factory. Journal of Electronic Commerce in Organizations 5, 2, (Apr.-June 2007), 123.
7. Kaptelinin, UMEA: Translating interaction histories into project contents. Proceedings of the SIGCHI Conference on Human factors in Computing Systems, 2003. 353360.
8. Mitra, A., and Gupta, A. Agile Systems with Reusable Patterns of Business Knowledge - a Component Based Approach. Artech House Press, MA, 2005.
9. Mitra, A., and Gupta, A. Creating Agile Business Systems with Reusable Knowledge. Cambridge University Press, Cambridge, England, 2006.
10. Mitra, A. and Gupta, A. Knowledge Reuse and Agile Processes: Catalysts for Innovation. IGI Global, Hershey, PA, 2008.
11. Rekimoto, J. Time-machine computing: a time-centric approach for the information environment. Proceedings of the 12th Annual ACM Symposium on User Interface Software and Technology. 1999, 4554.
12. Seshasai, S. and Gupta, A. The role of information resources in enabling the 24-Hour Knowledge Factory. Information Resources Management J. 20, 4 (2007) 105127.
13. Seshasai, S., Kumar, A., and Gupta, A. An integrated and collaborative framework for business design: A Knowledge Engineering Approach. Data and Knowledge Engineering 42,1, (Jan 2005), 157179.
14. Wiederhold, A., Tessler, S., Gupta, A., and Smith, D. B. The Valuation of Technology-Based Intellectual Property in Offshoring Decisions, Communications of the Association for Information Systems, forthcoming).
15. Gupta, A. Prescription for Change. The Wall Street Journal, October 20, 2008.
Figures
Figure 1. 24-Hour Knowledge Factory Concept will gradually apply to more categories of professional work
Figure 2. Time Difference is a Major Asset -Not an Impediment
Figure 3. All objects modified in last 24 hours can be viewed in context of evolutions.
©2009 ACM 0001-0782/09/0600 $10.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2009 ACM, Inc.
No entries found