Proven Practices for Effectively Offshoring IT Work

Senior executives dream of creating agile networks of global information technology service providers to lower costs, increase quality, implement seamless sunrise-to-sunrise production, reduce response time and disperse risks. Such agility allows organizations to enable flexible IT staffing while protecting business innovation.1 And indeed our research on 21 large U.S. offshore clients found that global outsourcing enables agile responses to business needs. (See “About the Research.”) For example, one U.S. financial services company has various joint ventures and fee-for-service relationships with 14 Indian suppliers. This network of suppliers enabled the company to adapt quickly to the immense surge in mortgage applications during the refinancing boom. As the refinancing boom subsided, the company was able to scale back resources immediately — all without affecting its domestic IT head count.

But agile IT networks require an immense amount of hands-on management. Our research found that U.S. clients micromanage their offshore suppliers to a much greater degree than they manage their domestic suppliers. The increased oversight is needed to mitigate higher risks, to build trust with new suppliers gradually and to coordinate delivery teams that are more remote and culturally diverse. Our research shows that, whereas U.S. clients require domestic-suppliers to submit status reports on one- to two-week cycles, many U.S. clients require offshore-suppliers to submit daily ones. While U.S. clients review résumés of domestic-supplier employees before they are assigned to teams, many U.S. clients personally interview every potential offshore-supplier team member to ensure that they have sufficient communication skills. And while U.S. clients trust domestic-supplier staff to ask for clarification when needed, they spend additional time questioning offshore-supplier staff to ensure understanding.

Micromanagement significantly increases transaction costs and can erode overall savings. Among our client companies, transaction costs for offshore projects neared 50% of contract value, compared to 5% to 10% for domestically outsourced projects. Much of the micromanagement is due to learning-curve effects and the labor-intensive nature of managing budding relationships with offshore suppliers. Clients want to know how to accelerate the learning curve, to mitigate risks effectively, to ensure cost savings while protecting quality and to work effectively with offshore suppliers. This article identifies emerging practices to address these issues. The practices were derived from interviews with 159 people, including clients, suppliers and advisers with substantial offshore outsourcing experiences. (Note that generic industry pseudonyms are used throughout this article to refer to the companies in the study. (See “About the Research” for the key.)

Practice 1. Escalate the learning curve with an aggressive, integrated program of pilot projects.

There is a typical client learning curve for offshore outsourcing.2 (See “The Offshore Outsourcing Learning Curve.”) During Phase I, managers become aware of offshore through marketing hype (“You’ll save 60% off your IT costs”) or irrational propaganda (“Software out-sourcing will hurt America’s supremacy”). Managers quickly learn about potential benefits, costs and risks by talking to peers and consultants and by reading research. Most managers initially begin offshore outsourcing with pilot projects to reduce costs on a few targeted projects (Phase II). As learning accumulates, managers move to Phase III when they exploit offshore for quality, as well as for cost reasons. More mature adopters in Phase IV use offshore outsourcing to enable corporate strategies, such as increasing business agility, bringing products to market faster and cheaper, financing new-product development, accessing new markets or creating new business.

The entire learning curve takes some organizations years to conquer. The slow pace begins in Phase II, where many companies we studied picked pilot projects that were too small and tested only a single supplier. False starts and switching suppliers kept Industrial Equipment (currently in Phase IV) in Phase II for three years. The process can be hastened, however, with an aggressive, integrated program of pilot projects. For example, a biotechnology company launched a series of 17 pilot projects with four Indian suppliers to test supplier capabilities fully, identify the most suitable project types and test the best contract mechanisms (fixed-price versus time-and-materials). In several instances, the client gave small pieces of the same project to two suppliers so that the project served as a control group for better supplier comparison. Within 18 months, the biotechnology company learned all it needed to move forward.

Practice 2. Select an offshore outsourcing destination based on business objectives.

Client organizations frequently selected offshore destinations by focusing solely on relative country advantage in terms of costs and risks. They often engaged a consulting firm to help identify the best location. But cost drivers and risk factors can rapidly shift, entrapping some clients to locations that no longer meet their needs. A better way to select destinations is to use a broader set of business criteria by considering the company’s strategic objectives and overall commitment to certain destinations.3 For example, one aerospace company selected Malaysia as its IT offshore destination because it hopes to sell planes in that country. The Malaysian government requires that some of the manufacturing be done in Malaysia, and the company’s IT presence will certainly help to meet that requirement. A hardware company selected China because it hopes to sell computers there. One U.S. client chose Canada because it wants suppliers in close physical proximity to its end customers for rapid deployment. Other participants selected offshore locations where they have existing manufacturing or research and development facilities. The existing facilities serve as a launch pad, with current employees serving as guides to the country, suppliers and culture.

Practice 3. Use offshore supplier competition to lower domestic supplier rates.

A truly global network of IT suppliers treats domestic and offshore suppliers as legitimate competitors. Clients with this view have integrated program management offices, rather than separate program management office based on supplier location. Clients from an integrated PMO request quotes from all suppliers instead of preselecting certain types of work for certain types of supplier. One consequence can be significant price reductions from domestic suppliers.

Consider the following example. Retail, a Fortune 100 company, has been actively engaged with 35 domestic suppliers since the 1990s. When domestic suppliers quoted very large prices for Y2K compliance in the mid-1990s, Retail began an engagement with a large Indian IT supplier. With the success of the Y2K project, Retail began to include more Indian suppliers on its preferred supplier list. That integration allowed a variety of suppliers to compete head-to-head on capabilities and project schedule in addition to cost. Retail could then compare the various value propositions from the domestic and offshore suppliers and expand its understanding of the offshore market. One consequence of the engagement was that it pressured its domestic suppliers to cut rates by 10% to 50%.

“We were paying about $100 for commodity-type coding [with domestic suppliers],” says the director of contract management at Retail. “The domestic suppliers saw the writing on the wall. We put out a bid to the approved list of domestic contractors and the current director of the PMO made it very clear that we were not going to pay those kinds of prices anymore. Our domestic prices dropped from about $100 to $80 per hour and some of the rates even dropped into the $50 range for some services.”

Practice 4. Diversify the supplier portfolio to minimize risk and maximize competition.

Concerning offshore suppliers, there are many choices. Some clients move offshore via one of their domestic suppliers such as EDS, IBM and Accenture. These established suppliers manage the offshore resources so the client doesn’t have to navigate through legal issues. Other clients, such as Financial Services 3, prefer to select Tier 1 offshore suppliers such as Wipro because of its maturity and stability. Other clients look for smaller niche suppliers with domain expertise. Still others select suppliers like Globalign that help clients find and engage offshore resources via large staff augmentation companies like Manpower.

The questions here are, which suppliers and how many suppliers? In a previous article in MIT Sloan Management Review,4 we covered the former question by offering a model of 12 supplier capabilities. That model is robust enough to compare the strengths and weaknesses of large domestic suppliers, Tier 1 offshore suppliers and offshore boutique suppliers.

Concerning the “number of suppliers,” most clients found that at least two offshore suppliers were needed to motivate continued performance. In contrast, reliance on a single supplier presents significant operational and strategic risks, including loss of market pressure on rates and a high concentration of intellectual property in one supplier. While maintaining engagements with multiple suppliers does entail additional transaction costs, the use of multiple suppliers successfully mitigates many risks associated with outsourcing.5

Industrial Equipment uses multisourcing to mitigate some of the risks associated with offshore outsourcing. It has active engagements with two large Indian suppliers and one small boutique supplier, which specializes in embedded software development. The three suppliers compete for additional work. According to the director of the software center of excellence for Industrial Equipment, “We don’t want to manage a whole bunch of vendors. However, we need the market pressure to make sure we are getting good prices. By letting the two large vendors and the small one bid on projects, and having the preferred list, we let them [the vendors] know that we are interested in enlarging the engagement, but we are also watchful of the costs.”

Practice 5. Allow business users to share in the benefits of offshoring to motivate adoption.

Among our cases, the use of offshore suppliers was either mandatory or voluntary. With mandates, senior executives (chief information officer or higher) required the IT department to meet minimum targets for the amount of work to be sent offshore. The idea was to “do more work” with a fixed IT budget. Business users were charged the same regardless of where the project was out-sourced. That put pressure on IT managers to persuade business users to allow their projects to be sent offshore. From the business users’ perspective, the additional overhead in dealing with offshore suppliers was a hard sell.

In contrast, voluntary use of offshore suppliers allowed business users to decide where to outsource projects. Business users shared the benefits of offshore by being charged less. The CIO of Financial Services 4, for example, allows strategic business units a choice for application development. The SBUs can outsource IT from three preferred offshore suppliers or from several domestic suppliers. Rates are lower with the offshore suppliers, but risks are lower with the domestic suppliers. The CIO believes the business unit managers should be the ones assessing the trade-offs.

Practice 6. Break projects into segments to protect intellectual property.

Recently, reports of sensitive data being stolen or purchased from Indian Business Process Out-sourcing suppliers have caused increased concern about the security of offshore data. In addition to common legal preventions outlined in contracts, there are a number of practices clients and suppliers use to protect intellectual property. Biotech, for example, required the offshore supplier to use Biotech’s secure R&D facilities in Bangalore for sensitive data access and updates.6 Many Indian suppliers create unique security areas that require special badges and a work-related need to gain access. Suppliers also create “clean desk” policies, which control what documents or computer screens can be viewed. While these practices are quite common, Industrial Equipment had a unique strategy for protecting intellectual property.

Managers from Industrial Equipment are very concerned about protection of IP because they outsource the development of strategic embedded software products. To mitigate the risk, they dispense work between the three suppliers mentioned above (two large and one boutique) to distribute the intellectual property effectively. They view their IP as a puzzle. By distributing small pieces among three suppliers, no one supplier can assemble the puzzle on its own. According to the manager of engineering: “We keep a very tight rein on where our IP is and who has it. We never let any one development team or any one vendor see too much at one time. We feel it would be impossible for our IP to be lost through offshore outsourcing.”

Utilizing multiple suppliers also helps Industrial Equipment limit the risk of a dependence on one supplier or lose the advantages of a competitive environment.

Practice 7. Ready the infrastructure.

Many of the clients we studied underestimated the difficulty in integrating offshore-supplier employees into the processes and work flows of their companies. Security concerns (access to systems and corporate data), human resource issues (modifying systems to use passport numbers instead of Social Security numbers) and the need to duplicate development and testing environments must be addressed prior to project launch. The PMO needs to take the lead in creating and improving the processes necessary to ease the “onboarding” of offshore suppliers. Some clients realized this too late.

“It really took us a long time to figure out how to make it [the onboarding process] run smoothly,” said the PMO director of one client company. “Since the suppliers needed access to systems from various business units and IT sectors, we had to cross organizational boundaries and create new protocols and rights profiles. However, without these processes, the suppliers sit idle waiting for us to ‘build a tunnel in the VPN [virtual private network].’ We should have had all these processes in place much earlier than we did.”

Practice 8. Understand how different contracts give suppliers different incentives.

There are a number of trade-offs between the two basic types of outsourcing agreements: “fixed-price” contracts and “time-and-materials” contracts.

Fixed-price Contracts.

In the offshore context, clients found that fixed-price contracts gave suppliers incentive to staff projects with their most productive people to increase their margins. But the downside was that the suppliers sometimes produced sloppy work (despite the productive staff) to meet unrealistic deadlines. As one client noted, “Suppliers have to make a reasonable margin to stay in business. You don’t want them to lose money because the worse their business gets, the worse your business gets.” The conclusion of most clients is that fixed-price contracts are most suited for projects with clear requirements so the suppliers can submit realistic bids.

Time-and-materials Contracts.

Clients found that time-and-materials contracts produced better work, but at considerable expense to the client. Suppliers were not pressured to take shortcuts in order to protect their profit margins. Their tendency was to follow all Software Engineering Institute’s Capability Maturity Model practices, or CMM, designed to protect quality because the client was paying for them. (CMM aims to foster processes that standardize, predict and continuously improve IT software development.7) However, overall costs can clearly be higher with time-and-materials contracts. In particular, the supplier is moti- vated to place new employees on the account because the client subsidizes the employee’s learning curve. Also, supplier employees who are unproductive take more hours to complete tasks, again reflected in the client’s bill. Some clients try to mitigate this risk by demanding to see résumés of supplier employees or by setting minimum years of experience. These practices place the client in the business of managing the supplier’s resources, which can increase transaction costs and create animosity between client and supplier.

One U.S.-client company experimented with both types of contracts. Satisfied with neither, it opted for a modified time-and-materials contract. It created a stratified pricing structure in which it paid as much as $10 more per hour for superior supplier employees. The client understood that better people are more productive, thus reducing overall costs.

Practices 9 through 11 address the process gap between U.S. clients and their offshore IT suppliers. Every study participant brought up the need to coordinate work processes, particularly with suppliers who are committed to the Software Engineering Institute’s CMM. While the Indian suppliers we studied were all certified at CMM level 4 or 5, the U.S. clients were usually ranked lower. At higher levels of certification, an immense amount of documentation is required. U.S. project managers had never been through such a rigorous process to define requirements.

At Biotech, for example, requirements definition is an informal process when it is done onshore. Project managers speak frequently with users who are usually located at campus headquarters. The user-feedback cycle is quick. In contrast, project managers, working on the offshore pilots, had to engage in many formal and planned communications with suppliers and users to create the required documents. One Biotech global team member said, “The overhead costs of documenting some of the projects exceeded the value of the deliverables.”

So what can be done to coordinate work more effectively with the supplier’s CMM processes? The following practices were used by participants.

Practice 9. Elevate your own organization’s CMM certification to close the process gap between you and your supplier.

Participants suggested that the best way to extract value from the supplier’s CMM processes is to become CMM certified yourself. The director of application development at Transportation had this to say: “A real problem we had was our CMM level 1.5 guys talking to the vendor’s level 5 guys. So together, we have worked out a plan with our vendor to help bring our CMM levels up. When we do, it will be a benefit to both of us; our specifications will be better and so the vendor can use them more efficiently.”

The outstanding issue is the level of certification required to work with suppliers effectively. The vice president at Financial Services 4 believes that clients need only to approach level 2 to extract value. The officer of IT services at Financial Services 3 believes client organizations need at least a level 3. Still other client organizations believe higher certification levels (at least CMM 4) are required to interact with suppliers.8

Practice 10. Bring in a CMM expert with no domain expertise to flush out ambiguities in requirements.

U.S. clients often complain that the requirements process is long and requires much expensive iteration. That is usually because the U.S. client doesn’t understand how the supplier will interpret the requirements. Some U.S. clients, for example, were surprised that supplier team members did not understand the concept of a mortgage. One U.S. client was surprised that the suppliers did not allow female name fields in the software to be altered unless a woman was recently married (as is the rule in India). In another example, the Indian supplier left-justified numeric data fields, causing the frustrated PMO director to lament: “Is this how ridiculously detailed my requirements are supposed to be?” To reduce the cycles during the requirements definition stage caused by misinterpretations, one CEO of an Indian supplier sought a unique solution. He brings in a CMM level 5 expert to the client site who purposefully has no domain knowledge. That enables him to identify ambiguities in the requirements documents that the offshore delivery team will likely have, thus reducing the number of iterations.

Practice 11. Negotiate “Flexible CMM.”

The project manager at Financial Services 1 noted, “You ask for one button to be moved and the supplier has to first do a 20-page impact analysis — we are paying for all this documentation we don’t need.” He is in the process of negotiating for exactly which documents Financial Services 1 will and will not pay. That will enable him to use only the CMM processes he perceives add the most significant value. While this practice is unique, a customized interface with each client could actually serve to increase the supplier’s costs, which may eventually result in much higher prices.

Some Indian suppliers are accepting, rather than fighting this practice by actively marketing the idea of “flexible CMM.” The managing director of a Tier 2 Indian supplier recognized the frustration and reluctance of U.S. clients to wade through the necessary steps for the supplier to maintain the integrity of its CMM processes. He says: “My clients are telling me, ‘You do what you have to do to pass your audits, but I can’t afford all of this documentation!’ So we have developed a flexible CMM model that maintains the processes necessary for high quality but keeps the customer-facing documentation and overhead to a minimum. Our customers have reacted favorably and our internal processes are still CMM 5.”

Practice 12. Factor in the use of an on-site engagement manager into the staffing models and ratios.

Due to the labor arbitrage available offshore, many clients attempt to recover the difference quickly between onshore rates (approximately $65 to $125 per hour) and offshore rates (approximately $25 per hour) by keeping few offshore suppliers on-site. However, due to the significant differences (cultural, technical, infrastructure, time zone, methodologies, language) between domestic and offshore suppliers, clients need to postpone the financial benefits by initially keeping more supplier staff onshore. Furthermore, the “right” people from the supplier must be selected to remain on-site.

While simply replacing staff may work with standardized and repeatable back-office tasks, IT development requires much greater integration and synergy between business users and the development teams. One key factor in ensuring successful integration is the use of an on-site engagement manager. Since the on-site engagement manager is familiar with the offshore supplier’s culture, accents, project management styles and internal workings of his or her own company, the use of an OEM will smooth the transition offshore and increase the quality of the deliverables.

Despite the significant costs associated with an OEM (domestic rates, travel, housing), clients new to offshore outsourcing need to utilize an OEM to lower the cognitive load on internal staff, increase project management transparency and protect internal employee morale. These benefits come from the fact that the employees chosen by the offshore supplier to be OEMs generally are far more experienced (technically and within the specialized verticals), have better English language skills, greater project management experience and are well equipped for client-facing tasks. These attributes will allow the client to shift the transaction costs (late-night conference calls, restating of requirements, time-zone bottlenecks and so on) to the OEM and relieve internal staff of these duties.

The use of an OEM will create a different staffing model and cost structure. For example, Biotech realized that in order to engage an Indian vendor efficiently, it had to staff the project differently. According to a technical leader: “If this project was to be staffed by domestic contractors, we would have just added two new contractors. However, since we were new to offshore, we priced in an OEM to interface between the business sponsors and the two offshore developers. We realized that all project cost savings were lost, but the OEM helped us improve our processes, interviewed and managed the developers and was responsible for status updates. We used the OEM to ensure cost savings on future projects.”

Practice 13. Give offshore suppliers domain-specific training to protect quality and lower development costs.

When managers from Industrial Equipment began to engage their Indian supplier, they realized they needed to manage the knowledge transfer process carefully. The supplier did not initially possess the domain expertise in the design and maintenance of embedded software, which differs substantially from traditional software. To bridge this gap, Industrial Equipment decided to give the supplier’s delivery team the same new-employee orientation sessions and training provided to internal employees. The content of this training included tools, methodologies and technologies used at Industrial Equipment as well as more traditional orientation subjects such as facility tours, introductions to peripheral departments and human resource issues. The technology and project training, which was conducted by the lead architects and project managers, dramatically increased Industrial Equipment’s transaction costs. However, these upfront transaction costs were needed to protect quality and lower development costs in the long run as the supplier became more knowledgeable and productive. Industrial Equipment made the supplier ensure that trained employees remained on the account for a certain period of time, or the supplier had to reimburse training costs.

Practice 14. Overlap onshore presence to facilitate supplier-to-supplier knowledge transfer.

This lesson is related to several other ones. Clients need to transfer domain knowledge (practice 13) to the supplier primarily by having the suppliers’ key liaisons, leads and managers on-site (practice 12), which increases costs. The knowledge transfer process between the client and these key on-site supplier leads can take weeks or even months. Furthermore, it is very difficult for these on-site supplier leads to manage remote teams. So some clients, like Insurance, end up bringing over much of the supplier’s delivery team. That is how on-site/offshore ratios can swell to a 70% to 30% split.

Industrial Equipment had a good solution to ensure knowledge transfer without overinflating the supplier’s on-site presence. It staged assignments so that the supplier’s next on-site project managers overlapped for three to six months with the previous supplier’s on-site managers. That way, the new on-site project managers were fully prepared to take charge when their predecessors departed. The predecessor then went back in India to train and manage the offshore delivery team, thus further disseminating knowledge. This overlap and ultimate transfer to offshore allowed the supplier to center the delivery effort offshore where rates are typically less than half of the onshore rates. In addition, there were other benefits. According to the engagement manager, “The overlap allows us to help ease the transition. We can share the stories and the history at a personal level. For example, there are ‘inside jokes’ that only the delivery teams would understand. We can transfer that ‘soft knowledge’ along with technical lessons learned about the creation of embedded software.”

Practice 15. Create balanced score-card metrics.

All participants identified the need for measures that consider costs, quality, timeliness and risk, but only participants from one company were fully satisfied with current assessment measures. Using a standardized activity measure, Industrial Equipment tracks in-house, domestic and offshore suppliers’ costs, quality and productivity. The data are captured by an in-house dashboard and analyzed monthly by management to monitor real development costs and trends. Managers learned that real savings from offshoring do not occur until after they have invested significant upfront training of every offshore developer and team leader. They also share this data with suppliers so that all parties understand the total cost trends.

Program management office managers use a variety of metrics to track performance of offshore work. (See “Balanced Metrics.”) The strategy and vision of an organization is important in determining not only which metrics it should track, but also the specific targets for those metrics. Indeed, a given metric can mean something very different from company to company and the targets for them can vary widely. For example, consider the metric “percent of supplier business,” which measures how much of the supplier’s revenues come from the client organization. One U.S. client wanted this number to be high, believing that a high percentage would motivate better service from the supplier. In another company, the client wants the metric to be low (less than 30%) because it didn’t want to be obligated for the supplier’s failure if it later chose to terminate the relationship.

Utilization rates are very important metrics to one U.S. client that uses a retainer model for offshore work. This company retains a certain number of the supplier employees each month for a fixed fee. If, for example, it pays a retainer for five employees for a certain month, it seeks to utilize at least 90% of the time for which it paid.

The comparative efficiency metric assesses the relative productivity of in-house employees to offshore employees. In most cases, in-house employees could produce more work (as measured in function points) than offshore suppliers in a given time period, but the offshore employees were significantly less expensive. Thus the target metric did not expect equal performance.

But the politics of measurement cannot be ignored. For example, one PMO manager said, “When I go ask our internal IT department to give me an estimate on how long it would take them to do this piece of work they say X, but my [offshore] supplier says Y. Who do I believe?”

OVERALL, OUR RESEARCH FOUND that offshore outsourcing can deliver on its promises, but it takes a tremendous amount of detailed management on both the client and the supplier sides to realize expected benefits. Clients that have used many of the 15 practices report that despite the additional management burden, offshore outsourcing of IT work is worthwhile. We have come a long way in four years,” says the client manager of Industrial Equipment. “The first time we did this [utilize offshore development teams], we thought we could ‘throw the requirements over the ocean’ and good code would come back. It was a terrible mistake, and looking back we really didn’t understand our own processes. We had to rethink our entire development process and analyze how we train our own people, how we manage the development process and how we actually develop code. Our second attempt is moving along much better.… Our vendors are not only providing a lower-cost talent pool, but they are helping us strategically. We keep looking for ways to increase the engagements. Our costs are down, productivity is up and the quality is as good as, if not better, than what we can do in-house.”