Captive/Dedicated Offshore Development Centers - are they really worth the effort?

I have been associated with the offshore space most of my career. I worked initially on the vendor side in delivery, account management and managing P&L for a region. I then moved into a US delivery organization initially to set up a global sourcing strategy and later on in to managing IT delivery using a mix of employee and third party.

I get a lot of questions on challenges faced by captives, JV's and dedicated development centers and thought this would be a good topic for this blog. This is based on my professional experience and research as well as interaction with colleagues in the industry. I have broken the topic into sections which I believe are core issues.

To keep things simple, when I refer to captives, I am generalizing the discussion of captives, dedicated (and owned) offshore development centers and joint ventures as they perform the same function, operate in the same way and have similar benefits and challenges. I have chosen to compare captives against the top players in the offshore industry. Captives are established by mid to large corporations and the alternative to setting up a captive is to work with a TCS, Cognizant, Infosys type organizations. In fact, as I explain below, captives often have to coexist and compete with these vendors.

Attracting Top Talent - The education system in India generates hundreds of thousands of engineers every year but the general feeling is that this is a case of quantity over quality. One only has to look at the schools the top industry players recruit at (by way of onsite campus events) and it is clear of where these companies believe they can find their top talents. Just to be clear - I am not talking about IIT's and IIM's but engineering schools. And companies like TCS, Cognizant, Infosys and Wipro hire anywhere between 10,000 to 50,000 fresh engineers every year so you bet they have perfected the process to make it efficient and worth their while. I do not mean to degrade or imply that the schools that are not able to attract these companies for campus recruitment are in some way inferior. But if companies want to find top talent - and in large numbers, they are likely to have better luck at some schools than others. There is intense competition to hire the top talent at each of these top rated schools between the big players itself, so it is reasonable to assume that a small company with a no name recognition would have a tough time competing with the biggies in attracting top talent from the top schools. You can still find good talent all over the country but it will take a lot more investment in time and money.

Training - You have to train new employees on process, methodology, standards, domain knowledge as well as update technology skills. Large organizations have dedicated infrastructure (e.g. Infosys's Mysore and TCS's Trivandrum facilities) that are in use pretty much all the time which makes the process efficient as well as cost effective. Smaller organizations cannot maintain dedicated infrastructure and have to pull out resources from other areas for training which impacts normal business.

Growth potential and Retention - Typical captives centers are built for a certain capacity and might take a few years to reach full capacity. However, it is extremely rare, in my experience that these centers continue to grow year after year after reaching steady state and this implies limited growth opportunities for the employees. (Almost) Every single software employee working in India is driven, ambitious, career oriented and looking to be a Project Manager in 4-5 years. This kind of career growth is simply not possible in a steady state captive which results in attrition. Employees who have reached their full productivity levels after 2-4 years are easy picking for competing organizations. This is frustrating for the management sitting in the western world - "It has taken us three years to get these employee to understand the domain - how can they just walk out like that"? Sooner or later this ends up increased compensation and other perks which don't really solve the problem but do alleviate the symptoms.

Costs - Captives, more often than not, follow the parent organizations policies. This would include swanky offices, business class travel, generous entertainment allowance, corporate events and more. I was surprised the first time I heard about a small captives team lunch at a 5 star restaurant but then realized this was the norm. However the biggest cost for the industry has always been labor and this is where the difference is remarkable particularly at senior levels where compensation levels are 2-4 times as much as what the top executives at the top players would make. The situation is even more extreme when organization depute executives from the west to manage these captives and provide top class accommodation and other benefits. The end result is that the per hour development cost of these captives turn out to be equal, if not higher, even though captives do not have sales and marketing costs!

Flexible workforce - It is difficult, if not impossible, for a captive to respond to short term requirements. If a project requires say 50 resources with a new skillset for building a new application, a captive cannot find their resources overnight and, more importantly, cannot release these resources after the project is completed. A larger organization can do this effectively. This limits the effectiveness of a captive to steady state efforts which generally turn out to maintenance and support work or where the long term technology strategy is very clear.

Captives v/s vendor competition - In most organization you will find situations where the captives are asked to compete with 3rd party vendors and this a great way to get a comparison and keep the captives in the game. However captives are cost centers and struggle to provide innovative pricing and upfront investment in emerging technologies not knowing if the projects will come through or not. More often than not, vendors invest a lot in onsite relationship building. As a result middle management decision makers on individual projects prefer to use vendors because of the relationships and track record. The use of captives, in spite of all good intention on part of the corporate organization, is seen as a forced directive and often considered a necessary evil.

Conclusion - Captives make sense when set up to protect IP, trade secrets or propriety knowledge. However the driving factor for the vast majority seems to reduce costs presumably driven by the high gross margins of the top players. The top players have economy of scale and control costs very closely. Although it is possible to replicate this model, it is certainly not easy. Organizations need to evaluate very carefully if these challenges can be overcome - and more importantly, if the effort involved in overcoming these challenges is really worth it.

Offshore Outsourcing at Worldspan

James O. Nelson

Director, Enterprise Technology Solutions

Worldspan

Ephraim R. McLean

Regents’ Professor and Smith Eminent Scholar’s Chair in IS

Georgia State University

Introduction

[Note: This paper was written in 2006 and it was a finalist in the Society of Information Management (SIM) national paper competition. It is still used today as a case study in graduate studies at Georgia State University.] 

The outsourcing of computer applications and computer-based activities is not new.  Having its roots in IBM’s Service Bureau Corporation (SBC) in the 1930s, the transfer of an internal process or activity to an outside third-party is nearly three-quarters of a century old.  It was greatly expanded by Ross Perot and his company, EDS, under the label of “facilities management” in the 1970s.  When Kathy Hudson transferred most of Kodak’s computer activities to IBM, DEC, and other companies in the late 1980s, the name given to this undertaking was “outsourcing,” the term that is used to this day. 

Following the success of the Kodak outsourcing contract, a number of computing vendors and service providers entered the outsourcing field.  However, most of these outsourcing contracts represented merely the transfer of existing corporate IT employees to the outsourcing company or “re-badging” them as it is sometimes called.  Often these employees sat at the same desk and did the same work as before; only the source of their paycheck was different. 

All of this changed in the mid-1990s, however, largely stimulated by the Y2K frenzy.  Lured by the appeal of low-cost IT talent in India and elsewhere, outsourcing vendors began shifting their IT activities from the U.S. to overseas.  Thus “offshore outsourcing” came into prominence.  The U.S. companies were soon joined by Indian firms like Tata, Wipro, Infosys, and others who launched their own outsourcing operations.

Outsourcing, in particular offshoring, does not come without its challenges and complexities, such as commercial agreements, immigration, staffing turnover, productivity, and many other factors   Companies are beginning to realize that there are indirect costs that must be taken into account.  Salaries in India are experiencing rapid escalation, annual staff turnover is over 30 % in some firms, communication across time zones is difficult, and cultural differences must always be taken into account.

All of these factors – both positive and negative – faced Worldspan in 2002 when corporate IT management began to explore ways whereby they could address the cost pressures they were facing. 

The Company

Worldspan is a leading independent provider of computer services to the airline and travel industry.  Its customers include Expedia, Orbitz, Priceline, and Delta Air Lines among others, with 1,860 employees and corporate revenues of $953 million in 2005.

 Worldspan’s history dates back to the late 1960s with the development of two airline-sponsored computerized reservation systems (CRS):  DATAS II at Delta Air Lines in 1968 and PARS at TWA in 1971.  Initially, these two CRSs were for internal use only; but in 1976 TWA began installing PARS in travel agencies, followed by Delta with DATAS II in 1982.  In 1986, Northwest Airlines purchased 50 percent of PARS; and in 1990, Delta, Northwest, and TWA agreed to combine DATAS II and PARS to form Worldspan, with headquarters in Atlanta, Georgia.  Worldspan’s data center in Atlanta became fully operational in 1993 and its systems unification effort was completed the following year. 

The company currently serves nearly 20,000 travel agencies in over 70 countries worldwide; and last year its computers handled 87.2 billion transactions, with an average of 39.6 million unique reservations on the system at any one time.  Its peak processing rate is 5,628 messages per second and it has 119 terabytes of online disk storage.  Because computer services are the business at Worldspan rather than supporting the business, the importance of providing accurate, reliable service to its customers is paramount.  In addition to providing airline systems, reservations, ticketing, and office automation services to domestic and international travel agencies, Worldspan operates the booking engine for over 50 percent of all Internet-based travel-agency bookings in the world. 

Why Outsourcing?

In early 2002, Worldspan, like many companies, was experiencing substantial cost pressures; and they began to explore the possibility of outsourcing some of their software development activities.  In particular, offshore outsourcing looked especially attractive to them.  Moreover, they soon realized that there are other potential benefits beyond just cost savings.  In particular, they anticipated achieving three major benefits through offshore outsourcing, all of which they have now realized:

• Acquiring a lower-cost, and yet highly skilled, professional workforce,
• Achieving “process uplift,” or improving the capabilities of the in-house professional staff by drawing on the expertise of the outsourcing vendors, and
• Gaining staffing flexibility, enabling the ramping up – and down – of the workforce in response to market conditions.

Although lower costs were the initial rationale for offshoring, Worldspan quickly realized they were acquiring the services of highly skilled computer professionals, in many instances equal to the skills of their existing workforce (although not always as experienced), all working for well established and respected vendors in India. Thus they were not making a compromise in quality in order to achieve cost savings.

This led to what Worldspan calls process uplift, or the ability to take advantage of the experience and skills of their vendors to improve the internal operations at Worldspan.  One example of this is CMMi certification from the Software Engineering Institute (SEI).  Worldspan is currently attempting to achieve CMMi certification for their own organization; and they have contracted with one of their vendor partners, who is at CMMi Level Five, to assist them in this effort.  This vendor is an authorized CMMi Assessment and Certification Company and they recently concluded a “gap analysis” of Worldspan’s internal development processes to help them move from Level One to Level Two this fall.

Also, as is true of many organizations, Worldspan, prior to outsourcing, was somewhat insular in their internal processes; they had no basis of comparing their in-house development processes to those of other companies.  Outsourcing allowed them to observe the processes of their vendors; and, in certain instances, they found ways to improve they own processes.  This cooperation on the part of the vendors to assist with this process uplift is so important to Worldspan that they have it written into the Master Services Agreements with each of their vendors. 

An example of one such internal improvement is in “billable hours.”  The work of each staff professional is divided into billable hours produced – productive work charged to a specific project – and infrastructure – training, vacations, and other non-billable work.  This ratio is a key indicator of the in-house staff productivity.  With the opportunity to observe the work patterns of their outsource vendors, Worldspan was able to adopt some of their “best practices” and improve their own internal productivity, thereby increasing the billable hours they could deliver without the need to increase the size of their staff.  Before outsourcing, Worldspan was averaging 1,650 billable hours per employee per year.  With the assistance of their vendor partners, this was increased to 1,780 hours per year – an increase of 7.9 percent. 

Lastly, staffing flexibility is essential, given the extreme economic swings that are occurring in the airline industry which comprises Worldspan’s main business.  The ability to expand, or shrink, the size of the vendors’ staff that are committed to Worldspan’s projects, while keeping their internal workforce constant, has proved to be extremely advantageous. The flexibility that this provides would alone justify the outsourcing arrangements that they have undertaken.

  

A Multi-Vendor Strategy

            Initially, Worldspan undertook a small project with IBM in India to gain experience with offshore outsourcing and to tap the global experience of IBM.  However, they soon realized that substantial benefits could be obtained by undertaking a multi-vendor strategy.  In 2002, they began a pilot project with Tata Consultancy Services (TCS), which concluded successfully two years later and led to a major contract with TCS in 2004.  More recently, because of the special expertise of InterGlobe Technologies in airline CRSs, they inked a third outsourcing contract.  Their present Master Service Agreements (i.e., their contracts) with these three vendors consist of the following personnel:  37 IBM staffers in Gurgaon and Pune, India; 137 TCS personnel in Chennai and Mumbai, India, with 38 more in Kansas City and Atlanta, and 22 in Montevideo, Uruguay (more about this later); and 48 InterGlobe staff in Gurgaon, India, and 15 more in Kansas City and Atlanta.  Worldspan now has global development centers (GDC) in six cities on three different continents.

           

           This multi-vendor strategy has a number of aspects that have proved to be highly beneficial to Worldspan.  They are as follows:

• Not placing all the eggs in one basket,
• A broader pool of resources upon which to draw,
• The natural competition that occurs among the vendors,
• The ability to obtain best-of-breed solutions,
• Built-in benchmarking, allowing comparisons among vendors,
• Through Master Service Agreements, projects can be multi-sourced among vendors, and
• With multiple vendors, a “near”shore/offshore strategy can be pursued.

Invariably, problems will arise in any client-vendor relationship; but if there is only one vendor, the client has few options when such problems occur.  With multiple vendors, however, if one vendor falters, the work can be shifted to one of the other vendors.  An example of the benefits of this multi-vendor strategy occurred in 2004 when the function of Passenger Name Record (PNR) Migration was outsourced to one of Worldspan’s offshore partners.  PNR migration involves the capture and movement of passenger-name records from one system to another and occurs when a travel agency, upon the expiration of a contract, moves its passenger-name records from one development center to another.  This occurs quite often and so Worldspan has historically maintained an in-house staff to handle these travel-agency PNR migrations.

            When the decision was made to outsourced this activity, the skill sets of each of Worldspan’s outsourcing partners were evaluated and a contract was let to the firm that appeared to have the necessary analysis and programming skills.  However, it soon became apparent that, although they had the technical skills, they did not have the specific development tools and, more importantly, the domain knowledge concerning PNR migration; and it would take too long for them to come up to speed with this requisite knowledge.  This delay was unacceptable because it severely affected Worldspan’s ability to attract new business, since PNR migration is an important part of travel-agency conversions.

            In spite of attempts on the part of the vendor to catch up, the delays made it imperative for Worldspan to move to another vendor.  Luckily, one of their other outsourcing partners was already doing some PNR migration work for another company and therefore had both the technical skills and the domain knowledge to begin the work immediately.  As an aside, it should be noted that PNR migration work, although an essential component of servicing travel agencies, is not a source of competitive advantage; it is more of a commodity service and thus the fact that the perspective vendor was doing this work for a competitor was not a source of concern.        

           

            Worldspan asked this vendor for a bid on the work and received one that it was competitively priced, similar to the contract that was in place with the first vendor.  Because of a non-performance clause in this first contract, there were no costs in canceling it and so Worldspan gave a contract to their other vendor partner.  The transfer of work was accomplished with a minimum of delay; with a corresponding minimal impact on Worldspan’s customers. As a side effect, the offending vendor’s performance on its other projects for Worldspan improved markedly.  Had they not had these well-established vendor relations, Worldspan would have been forced to bring the project in-house, resulting in considerable disruptions.  

This illustrates the benefit of having a healthy competition among vendors and the value of benchmarking their performance in the awarding of new and follow-on projects.  In order to ensure such competition among vendors, the Master Service Agreements are also written in such a way as to allow multi-sourcing of individual projects among two or more vendors, further increasing Worldspan’s flexibility in project management.

      

            Most vendors, especially the larger ones, will claim that they have all the resources that are needed for any project.  In reality, some companies are better at some things than others and their skill sets reflect this expertise.  By using multiple vendors, Worldspan has been able to tap into these specialized skills, drawing on a broader pool of resources.  For instance, as discussed above, Worldspan entered into a contract with InterGlobe Technologies in order to tap its special expertise with airline reservation systems. 

Nearshoring

            The last item in the list above introduces a concept that Worldspan has found quite valuable to them, that of “near”shoring.  As discussed above, the first decision that companies must make is whether or not to outsource at all; and, if so, how much and will it be onsite – re-badging – or offsite?  If it is to be offsite, will it be onshore or offshore?  This has typically been the choice pattern.  In the case of Worldspan, they currently split their work as follows:

            Onsite

·        Technical leaders,

·        Business analysts,

·        System design, and

·        Project management.

Offsite

·        Application development,

·        Testing,

·        Maintenance,

·        Technical support, and

·        Business process support.

In making this division of labor, Worldspan reviewed their internal development methodology, or Framework as they call it, to determine where in the systems development life cycle it would be most appropriate to consider outsourcing.  They felt that they must keep the business analysts and application designers close to the business and product strategy organizations in order to capture the requirements close to the source.  As long as the in-house analysts and designers do a superior job of defining and capturing the requirements, then Worldspan feels that they can more readily move the subsequent application development activity offsite, whether to another Worldspan location or to a third-party vendor.

            Worldspan has maintained remote development centers in both Kansas City and Ft. Lauderdale for some time, and thus has experience in having development work done remotely.  By keeping analysis, design, and overall project management onsite,

application development, testing, and maintenance are attractive candidates for transfer to an offsite location.  By using virtual classrooms, VoIP, and instant messaging to connect to remote locations, the coordinating problems are minimized.  Each site is integrated into Worldspan’s communication system so a development manager across the country or across the globe appears to be as close as across the hall.  Five-digit fast dialing is available to all those locations with which Worldspan does business.

            Worldspan’s clerical and order-fulfillment work, what they call Business Process Support, is an ideal candidate for outsourcing.  It tends to be repetitive and follows set procedures which are well documented.

            So, once the decision was made as to what should be outsourced, the next step was the decision to go offshore to India to take advantage of the cost/quality opportunities.  But, in the case of TCS, this vendor had locations all over the world, not just in India. In particular, they have established development centers in a number of western hemisphere locations, including Montevideo, Uruguay. This provided Worldspan with the opportunity to leverage their already-existing relationship with TCS and establish a “near”shore presence in this location. The Uruguayan center offers a number of advantages:

• Business hours are similar to those in Atlanta,
• A follow-the-sun development cycle,
• Geographical diversification,
• Reduced costs through a better staffing mix, and
• Fewer immigration and visa processing problems.

Having a development center in Uruguay in a similar time zone as Worldspan’s headquarters in Atlanta offers a distinct advantage over their outsource vendors in India, which are ten hours out of phase with Atlanta.  On the other hand, this allows a follow-the-sun development cycle, where TCS in Chennai, India, can work on a project during the day; pass it onto their center in Montevideo, Uruguay, where they work on it during their day; and then they pass it back to Chennai, to start another day – effectively creating a ”round-the-clock” development cycle. 

Geographical diversification provides protection against possible disruptions that might occur at a location, and also allows for the creation of special staffing expertise at a particular site.  Lastly, because of the lighter demand on the U.S. Consular Service in Uruguay for visas and permits to work in the U.S. – as compared to the huge demand for them in India, often causing delays of months or even years – these documents can be obtained in a relatively short period of time, adding to the attractiveness of Uruguay for the movement of staff to and from Uruguay and the United States

Risk Mitigation

            Another aspect of Worldspan’s outsourcing strategy is its attention to risk mitigation (see the table below).  Two key aspects of this risk mitigation strategy are governance and measurement.  Worldspan has created a Global Resource Office (GRO) charged with overseeing the entire outsourcing program.  The GRO is a small centralized group that directly manages the Master Service Agreements and the draft Statements of Work, oversees security at the global development centers, and administers the Service Level Agreements that are in place with each of their vendors. This GRO produces the metrics, benchmarks, time management reports, dashboards, and so forth that are central to Worldspan’s staying on top of their outsourcing relationships. The following metrics are produced by the GRO on a regular basis:

• Performance against budget and against plan
• Average hourly billing rate, combining in-house and outsourced rates
• In-house versus outsourced ratios
• Staffing and skill levels by FTE
• Total hours delivered by outsource vendors
• Vendor hours delivered by category, for both productive and non-productive work
• Vendor productivity rates
• Staff augmentation versus fixed-price expenses

These measurements and resulting reports are reviewed weekly with the vendor representatives.  Many of these metrics are integral parts of the Service Level Agreements and thus directly measure the performance of the vendors.  In addition, the vendors conduct monthly project reviews with the Worldspan department heads that are sponsoring the work in order to review the performance and challenges from the prior month.  From these weekly vendor reviews, to the detailed monthly project reviews, Worldspan is able to take timely corrective actions in order to ensure improved quality and the overall performance of the outsourcing activities.

Risk Mitigation Strategies
7X24 development	·Share work among existing GDC around           the world ·Establish some services nearshore
Productivity enhancements	·Use comprehensive metrics in the          Statements of Work (SOW) ·Have senior-level project leaders onsite ·Shift time-and-material contracts to fixed-          price contracts
Cultural Integration	·Promote Indian-American cultural          awareness and sensitivity ·Conduct subject-matter-expert (SME)          training onsite at GDCs
Billing and Invoicing	·Have vendors enter their time directly into              the Worldspan reporting system ·Approve these time reports weekly ·Reconcile these time reports against              vendor invoices monthly
Knowledge Acquisition	·Introduce a technical “Knowledge Portal”          for the sharing of information ·Establish benchmarks for vendors for          acceptable levels of turnover
Security	·List all vendor personnel in the corporate          directory ·Have dedicated secure-entry GDCs

           

            Some of these risk mitigation strategies, such as 7X24 development, have already been discussed above, as have some of the efforts that have been undertaken to improve productivity. Having senior-level project leaders onsite has been particularly helpful in accomplishing Worldspan’s process uplift initiatives, as has been the establishment of the technical “Knowledge Portal” for the sharing of best practices.

            The cultural issues, sometimes troublesome in some offshore outsourcing arrangements, have been faced head-on, with personnel from both sides visiting their opposite numbers, eating their food, partaking in their festivals and events, and thereby developing an awareness and sensitivity of each other’s cultures, 

Another issue, often the subject of disagreement, is service billing.  Worldspan has addressed this by having all of their vendors post their time reports directly into Worldspan’s system, where they are approved on a weekly basis and reconciled with invoices on a monthly basis.    

            One risk factor that all companies doing business with Indian outsourcing firms face is turnover.  Given the rapidly escalating demand for services and the limited supply of qualified professionals in India, salaries are being bid up and people are switching jobs for more money.  Turnover rates of 30 percent per year are not uncommon.  To protect themselves against this, Worldspan has written into each of its Master Service Agreements with its vendors that personnel assigned to their projects cannot be reassigned without the approval of Worldspan.  Naturally, if personnel leave the vendor’s employment, Worldspan cannot control this; but, if the level of such turnover on their projects become excessive (as defined in the MSA), the vendor is obligated to pay a penalty to Worldspan.  Thus the vendors have a decided incentive to keep their employees happy so that they do not leave.

            On governance, Worldspan has gone to great lengths, with the assistance of an outside law firm that specializes in outsourcing contracts, to craft detailed Master Service Agreements, Statements of Work, and Service Level Agreements with each their vendors.  By having such agreements in place, with deliverables, timely responses, and staff augmentations clearly spelled out, the areas of dispute have been greatly reduced. 

Another related effort in governance is the move from time-and-material costing to fixed-price contracts.  The first step in this is staff augmentation, where the vendor “augments” the Worldspan staff by doing application development and testing on systems designed by Worldspan.  While easier to implement, staff augmentation places a significant oversight burden on Worldspan’s management.  The next step is out-tasking, where a much larger percentage of the work is tasked to the vendor, from detailed design through to the delivery schedule, including TQM, training, and project management.  Worldspan’s focus today is in shifting its staff augmentation work to the second step of out-tasking.  The final step on the road to totally fixed-priced contracts is what Worldspan calls portfolio.  This approach bundles multiple activities into an agreement for not just application development, but also for additional services, including maintenance and production support, into a more holistic product life cycle plan, thus vastly simplifying the governance issue.

Lessons Learned

Worldspan has identified a number of factors that have contributed to its outsourcing success; although it is still a “work in progress,” and they are still learning.  Many of these critical success factors have already been discussed above; they are summarized here.

Multi-vendor strategy.  The decision early in their outsourcing efforts to pursue a multi-vendor strategy has been a key success factor for Worldspan. For the many reasons outlined above, this has reduced their risk exposure, leveraged the natural competition among their outsourcing vendors, and provided a much broader talent base than any one firm could have provided

Process uplift.  By being open to new ideas and tapping the expertise of their vendor partners, Worldspan has been able to learn from them and improve their own internal staff productivity.

Governance.  Through the creation of the Global Resource Office, Worldspan has a central, but adaptive, sourcing office. This office, coupled with the senior-level project leaders who are onsite, the frequent visits to vendor sites around the world, and the constant monitoring and fine-tuning of the relationships, has allowed Worldspan to exercise careful control over its outsourcing activities.

            Measurement.  It goes without saying that none of the above would be possible to manage and maintain without comprehensive measurements.  Fourteen major reports for management are produced on an on-line basis, backed up by more detailed reports for technical managers.

            Culture, mutual respect, and trust.  “Partnership” is probably an over-used and mis-used word, but it aptly describes the relationships that Worldspan has built with its vendor partners.  This cultural understanding and trust goes a long way in easing the many differences that naturally arise.

            Legal guidance.  When Worldspan began negotiating their first outsourcing contract, they were quite inexperienced and thereby at a distinct disadvantage in dealing with the outsourcing vendor, that had considerable experience in such negotiations.  Worldspan therefore retained the services of a law firm that had extensive experience in outsourcing contracts, and this proved invaluable. However, they made sure that the resultant contract was business-driven, not driven by the lawyers.

            Top-level support.  Although last, this factor is by far the most important of all these listed   Executive support at the “C” level was crucial, especially in the beginning, where it was clear that costs would climb before the savings would kick in.  Also, the formation of an executive steering committee helped clear the way for the road blocks that inevitably occur.

The Results

  

Worldspan began its offshore outsourcing in 2002, with the first major contract let in 2004; thus they have less than two years of experience.  Yet, according to Ron Richards, Worldspan’s Corporate Director of Internal Auditing, measurable gains have already begun to be realized.  Using Worldspan’s internal productivity metric of “billable hours produced,” the combined output of the in-house staff and their outsourcing partners has improved by 7.9% in the last year. 

Using 2003 as a baseline, Worldspan's application development staff budget has decreased by more than 20% as a result of its outsourcing program.  In 2003, total staff costs, both in-house and outsourced, were $93.4 million.  In 2004, these costs were $84.2 million; in 2005 they were $82.6 million; and this year they are projected to be $71.3 million, for a cumulative savings of $22.1 million.  These figures include the non-staff expenses related to outsourcing of $45K for hardware and software, $147K for networks and telecommunications, and $182K for travel expenses.

As reported by Kim Nancarrow, the Director of Financial Planning and Analysis, these savings represent employee-related cost improvements that are continuing into 2006 and beyond.  But equally important to the cost savings are the improvements in Worldspan’s internal development processes, further enhancing its dominant position in the travel industry.

Portfolio Assessment for Outsourcing Contracts

People say that change happens for one of two reasons: evolutionary or revolutionary. When considering outsourcing information technology (IT), change rarely happens due to evolutionary means, since it is not a natural occurrence to intentionally decouple IT systems and move components to another entity at a different location. In these days of globalization outsourcing usually means turning over aspects of your application portfolio to a company based outside of your own country and operating in diverse geographic locations spread across multiple time zones.

IT change is more often revolutionary and the decision to outsource is made by a senior executive committee or even by a non-IT group, such as the Finance department. So, the question presented to the IT group is not “Do we outsource?” but more “What to outsource?”

It’s widely known that certain aspects of the software development life cycle (SDLC) are better suited for outsourcing than others. As an example, interacting with the end users and eliciting requirements is best managed by analyst that are part of the company owning the customer relationship. On the other hand coding and testing are easily adopted to turning over to a third party who may have a larger pool of resources from which to draw and may even have better processes support application development. Handing over a legacy portfolio to a qualified third party to ‘moth ball’ or simply ‘keep the lights on’ is fundamental to most large outsourcing vendors, and in doing so allows the outsourcer to focus on new products or a technology refresh.

Other factors weigh in on decision regarding what to outsource, such as product life cycle, budget, domain knowledge, and overall readiness to support an outsourcing engagement. Within this in mind the process in determining what to outsource should not be taken lightly. Recognizing there is no single parameter or component that is a complete indicator of which portfolios are right for outsourcing, it is generally an entire set of factors that combined give indications when one portfolio may be more appropriate than another for outsourcing.


When evaluating a portfolio for outsourcing potential, I suggest a balanced approach in assessing the portfolio, so that no one factor overshadows the other factors or adversely influences decisions. My experience has shown that there are three categories of parameters that must be considered:

• Business contribution of the portfolio
• Skills required for development and support
• Platform characteristics on which the portfolio operates

There is a simple but dynamic process for evaluating portfolios and assessing their potential for successful outsourcing. Based on years of experience in managing millions of dollars in outsourcing engagements and overseeing hundreds of projects and statements of work (SOW), I have found a set of criteria that are comprehensive, yet equally aligned with the three categories. A simple questionnaire that scores responses and can be utilized to rank potfolios relative to each potfolio's adaptability as determined by proven industry best practices.

Business factors

1. Business Continuity, what is the risk to going business of a failure in this portfolio?

2. Competitive Advantage, does this portfolio provide an advantage over similar product from competitors?

3. Investment Budget, is the annual portfolio budget for development and support greater than the minimum threshold to justify the ramp-up costs?

4. Relationship Complexity, how complex is the contributor arrangements, such as the number of vendors or stakeholders?

Skills Related

5. Domain Dependencies, what, if any, domain knowledge dependencies are there for successful sourcing by a third party?

6. Skills Availability, how available are the skills in the open market that are required supporting all development and operational needs?

7. Readiness, are SMEs trained and are tools in place to adequately support the sourcing ramp-up?

8. Knowledge Transfer, what is the length of time it takes an average technician to become productive on the applications?

Platform characteristics

9. Product Life Cycle, where is the portfolio (or product) in its life cycle?

10. Platform Stability, how complex are the systems and tools on which the applications will run?

11. Maintenance, what percent of the portfolio budget is allocated to maintenance and production support?

12. Instances, are there multiple instances or derivatives

As I have indicated not every portfolio is easily adapted to outsourcing. Once vendors have been vetted and contracts signed, then changing vendors or brining the development back in-house can be even a larger initiative than the original outsourcing due to ramp-down and knowledge transfer. A focused assessment of portfolios prior to outsourcing can pay huge dividends in the end.