Dynamics and economics of offshore software development model
a project management perspective
What is offshore outsourcing? The process of developing any part of your software project with a company or developer that has a base in another country is considered offshore outsourcing. This process began as a cost-saving concept; however, it has now developed into a structured business model, rendering a cost-effective solution, efficiency and quality service. According to the International Data Corporation, the offshore outsourcing industry has doubled from $5.5 billion in 2000 to over $10 billion today. By 2005 the projected spending for this outsourcing practice will be over $17 billion. As the business of outsourcing evolved, so to did the development of management models, tools to assess the process and technology to support the offshore outsourcing initiative.
There are several factors that drive offshore outsourcing. One factor may be that a company needs to reduce or control their operating costs. It would be better for a company to outsource than to build upon its infrastructure to complete a project, thus not affecting the overhead. Another factor for outsourcing offshore would be the gained access to world-class capabilities that would, otherwise, cost a fortune to create in-house. Along these same lines, most companies have insufficient resources to complete a sizeable software development project in house. Sometimes companies realize that a development project would be too time consuming to pull their own resources into a project, so they utilize offshore development companies with the experience, knowledge and resources to complete the project. As well, they may consider specifically off shoring their development project with a company that is on the opposite time spectrum so that projects could be continuously worked on throughout the workweek. One of the biggest factors in offshore software development is the “shared risk” concept. A company that partners with a technology team will share risk in both business and technology, thus alleviating some of the pressure from development of the project.
Once a company decides that their needs justify offshore outsourcing, the next step would be to decide what delivery model would best fit the company infrastructure. There are six modes of delivery; fully onsite, mostly onsite and partly offshore, partly onsite and mostly offshore, fully offsite, fully offshore, or globally distributed and supported. In order to make this decision there are certain factors that a company should take into account. What type of engagement will be used in a given situation? What will be the estimated project effort? Where will this offshore partner be located and what is the company's perceived cultural issues? Also, what is expected from the direct contacts, which will highly depend on how detailed the specifications are.
Fully onsite is when the whole process, starting from information gathering to implementation is done at the host company's premises by an offshore partner. Based on the needs and requirements of the host company, the design, development and test teams are deployed for a short time frame at the host company's location. This is suitable if the project needs a specific resource type of post-deployment, maintenance, support and follow-up activities. This model is helpful for those projects that are mission-critical, require proper and constant attention and, because of the host company's comfort level, needs everything to be done at the host company's location. Thus, this model is ideal for:
- - Requirements not being defined at length, forcing dynamic changes in deliverables or requirements, open-ended and iterative nature of project scope
- - Tough and rigid deadlines
- - Constant need of support and direct interaction with client required
- - Mission critical projects and product engineering-related services
- - Projects that are highly secured and confidential
In this approach, all the project phases - discovery, definitions, executions and integration are done onsite.
In the fully offsite model, the consulting company will have its office near the host company location which will act as a coordinating centre, but it also gives support to the onsite team and the offshore development activities at the offshore centre. This outsourcing model is appropriate for short-term business process projects. Also when the requirements are defined beforehand and there will be no frequent changes in the project, and when the host company does not have the extra capacity to facilitate another person onsite.
In the fully offshore model all the project-related activities; from the initial study to testing is done at the consulting company's premises. The consulting company will not have any presence at the host company's location, but the client will interact directly with the offshore team. This model works best when the project plan is well defined and the development team has a clear understanding of host company's requirements. The team members at the offshore location interact with the client through various communication modes, such as telephone, fax, and email. This model has proven to be effective in terms of high quality service with low labour cost, effective utilization of a time zone (the 24x7 service), the availability of multi-technology skills and 30% to 50% reduction in project cost. However, the risks associated with this model are the communication gap between the host company and the offshore consulting company and the host company's requirements may not be captured in real terms.
There are two variations of the offsite/offshore delivery model:
- Mostly offshore partly offsite;
- Mostly offsite and partly offshore.
Both these models of outsourcing provide value to the host company. The consulting company will have its office close to the host company's location and offshore outsourcing facilities at some other geographical region, which enables seamless transition and execution of projects without compromising quality and control for the host company to manage. The offsite outsourcing facility serves as a medium of communication between the host company and the offshore outsourcing development team. Both the offsite and offshore outsourcing facility should be connected with highly secured network facilities for seamless communication with the respective host company. This connectivity set up ensures failure-free, fast and secure access for the host company to monitor the progress of development and support activity at both the offsite and offshore outsourcing locations. In this model the consulting company should have an efficient backup facility for regular project backups and disaster recovery procedures. Generally, the distribution of work between the two outsourcing facilities is as follows depending on the project:
|Offsite outsourcing:||Offshore outsourcing:|
|Requirement Gathering & Analysis||Detailed Design|
|Project Management Planning||System Testing|
|Acceptance Testing||Quality Assurance|
|Delivery||Support & Maintenance|
The onsite/offshore delivery model gives an opportunity for the host company to have direct communication with the outsourced consulting company through the onsite centre and at the same time gives them the chance of enjoying the benefits of offshore outsourcing. Ideally, 20%-30% of work is done onsite whereas 70%-80% is outsourced offshore depending upon how critical the project is. Thus, the whole gamut of benefits for such outsourcing can be summarized as:
- - Cost-effective in terms of time, money & communication;
- - Faster, better and easier;
- - 24x7 work cycles;
- - Availability of multiple skill-sets;
- - Low cost for resources;
- - On-time and high quality service;
- - High security set-up;
- - Direct interaction between host company and outsourced consulting company.
Hence, under this delivery model, the distribution of activities is as follows:
|Onsite outsourcing:||Offshore outsourcing:|
|Initial Study||Knowledge Transfer|
|Requirement Analysis||Detailed Design|
|Project Management Planning||Milestone-based Development|
|High-level Design||Integration Testing|
|Acceptance Testing||Application Management|
The final model is the global distributed and supported delivery, which is similar to the onsite/offshore model. The key to this model is a global support infrastructure in different geographical locations either of the consulting company or one or more of their partners. Each of these facilities will act as an offsite development team for one part of a given project. The outsourcing consulting company ensures a well-planned mechanism under which the outsourcing project resources are distributed physically at multiple locations. This kind of outsourcing methodology ensures:
- - Low cost benefits;
- - High quality outsourcing service;
- - Resilient to disasters;
- - 24x7 support;
- - Lower outsourcing project risk;
- - Improved delivery time;
- - High security set up.
This outsourcing delivery model aims to productively spread out and manage engagements and resources across multiple global locations allowing the consulting company to better respond to dynamic host company requirements around the globe at unsurpassed levels of speed, quality and value. The biggest advantage of this outsourcing model is that it saves the host company from investing in a huge team of professionals. It conveniently adapts to the changing requirements and, if there is a sudden need for a huge team, the outsourcing consulting company can immediately supply them and later relocate them to other locations conveniently. The global delivery outsourcing model is the appropriate mix of onsite and offshore outsourcing delivery modes that results in quicker deliveries, improved end-user requirements, and excellent cost-effectiveness. Moreover, if any disaster strikes in any one of the locations, the outsourcing service provider can immediately pull up resources from other locations so that there is no interruption in business process. The distribution of outsourcing activities across the onsite and offshore locations may vary according to the demands of the project
Dynamics of Offshore Outsourcing
After careful consideration on deciding which delivery model to utilize, the host company now needs to consider the dynamics of outsourcing these projects offshore. What we mean by “dynamics” is what are the key issues or critical areas in off shoring. A key dynamic is the perception of the agreement or contract. These critical issues will affect the success of the project. An offshore deal should address these critical components:
- - Defining clearly what the host company wants;
- - Objective ways to measure the deliverables;
- - Flexible agreements;
- - Offshore partners understanding of the project objectives and goals in line with host company's short term and long term objectives;
- - Managing the relationship between the host company and consulting company.
There are a lot of similar guidelines and critical areas in offshore outsourcing no matter which delivery model is utilized. Critical issues for a successful outsourcing program, in a recent survey (Survey of Current and Potential Outsourcing End users, The Out Sourcing Institute Membership, 1998) have been identified as:
- - Understanding company goals and objectives;
- - A strategic vision and plan;
- - Selecting the right vendor;
- - Ongoing management of the relationship;
- - A properly structured contract;
- - Open communication;
- - Senior executive support and involvement;
- - Careful attention to personal issues;
- - Short-term financial justification.
From the above lists, we believe the following are key dynamics of offshore outsourcing projects:
- - Consulting partner's understanding of the host company's goals and objectives
- The consulting company and their team should understand why this project is important to the host company and how it will fit in the overall scheme.
- Host Company should take painstaking measures to make sure that this is understood.
- - Relationship management (host and partner)
- Should be treated as a necessary program as inducting a new employee
- Accessing stakeholder requirement is the first part and having open channels of communication is vital. Everyone should be involved and this should be a managed process.
- - Senior executive (from host company) support
- Offshore outsourcing initiatives are strategic and should be supported by the upper echelons of the host company.
- Goals and objectives of similar initiatives should be articulated at a senior level and communicated within the company on the benefits of the program.
- - Sound and flexible contract
- Negotiating for reasonable contract
- Service level agreements (SLA) should be included to help manage the relationship, and to identify responsibilities. Focus should be on results (measurable, quantifiable and comparable against the pre-defined criteria).
- The contract should be flexible and should build in post contract issue identification and have room for further negotiations.
Economics of Offshore Outsourcing
Another area of offshore outsourcing model that should be analysed before going forward is the economics of this type of delivery. We are not reaffirming the considerable cost effectiveness, which has been studied and documented. There are actual key economic indicators relevant to projects and identified by the International Software Benchmarking Standards Group (ISBSG). We chose three of the many economic indicators for further study and for our analysis. These three indicators are the project delivery rate, the speed of delivery and the project duration. These were selected for the following reasons; they affect the overall economics of the project, there are measurable and comparable data available from the ISBSG, for a variety of projects and delivery models, as well, from our experience have observed that these represent economic and measurable value in offshore outsourcing projects.
Below are definitions and general information on the three economic indicators taken from a public domain provided by ISBSG.
Project Delivery Rate, PDR, is a measure of the hours it takes to deliver a Function Point. A low PDR means that fewer hours were needed to deliver a function point. It is defined as Project Work Effort (measured in hours), over Functional Size of the delivered software (measured in function points). It is expressed as hours/FP. ISBSG publications report the PDR's for the projects in the repository.
These PDR figures can be used for project estimation or to benchmark project performance against similar projects in the repository. There are a great variety of projects in the repository. The PDR varies considerably from one group of projects to another, so it is important that you compare “apples with apples” when benchmarking or estimating using PDR.
Speed of Delivery is the measure of the number of function points delivered in an elapsed month.
The Practical Project Estimation Toolkit provides a table of regression equations that one can use to obtain an indicative estimate of speed of delivery for a project. Equations are provided for by ISBSG.
- Development platform
- Development language
- Development language/platform combination
Project Duration is the measure of the elapsed months that a project is completed in. As with most aspects of the software development process, software duration has many influencing factors.
Functional Size and Duration Relationships - One can estimate project duration using estimation of software size, software development life cycle phase profile, (including overlap between phases and activities), staff availability and other factors. For a “ball park” or indicative estimate of duration, several “rules of thumb” equations, or quick estimating techniques are available.
Relationship between dynamics and Economics
Within the rest of this paper we will discuss all of the above issues and how they affected eight different projects. We collected data from projects with different development platforms, language, project size, team size and different business perspectives. As well, each project involved a different delivery model. We noted the key dynamics for each project and scaled them. Also, we calculated the economic indicators for each project. Then, by comparing the dynamics and economic factors we discuss what the relationship is between the critical areas in offshore projects and its economic indicators. After that initial comparison, we compared those findings against the ISBSG data repository of international project data and created a framework for other project managers to utilize when organizing their outsourcing projects. This will guide project managers on what to do and not do in offshore engagements, and, equally important, help them identify the project risks.
Below is collected data from eight different projects done during the last couple of years to study the relationships between dynamic issues and economic indicators identified earlier. Again, these projects selected for study are of different sizes, involving varying team sizes, and developed using different engagement models. But all the projects were executed by offshore outsourcing. There are factors that are kept common for all the projects to understand the pattern and relationship, for example the use of fourth generation languages (4GL's). All the projects followed a defined and documented process.
Throughout the study we used Data Collection Questionnaire - New Development, Redevelopment or Enhancement v.5.8 (IFPUG/NESMA) as a tool to collect and organize information on our projects.
The following table gives an overview of the projects selected for study:
|Project||Engagement Model||Business area||Type of project||Technology, tools, Operating Systems|
|1||A||Mostly offshore||Health Care||New Development||Web Based, Enterprise Java/Oracle/Windows|
|2||B||Mostly onsite||Health Care Communication||New Development||Web Based, Enterprise Java/Oracle/Windows|
|3||C||Mostly offshore||Transportation||Re-Development||Web Based, Enterprise Java/Oracle/Unix|
|5||E||Mostly offsite||Public Health||New Development||Web Based, Enterprise Java/Oracle/Windows|
|6||F||Offsite||Health Care||New Development||VB/SQL/Windows|
|7||G||Onsite (by offshore team)||Insurance||Enhancement||VB/SQL/Windows|
|8||H||Onsite/offshore||Data Management Tool||New Development||VB/SQL/Windows|
Table 1: Project Description
To represent the dynamic factors of the project, we used the scale measurement. The persons involved in the projects were asked to scale the factors (to the scale of 1-5) based on their understanding, involvement and knowledge of the project. The representatives of both the host and partner companies were involved at this level of the study. The aggregate of the observations are given below in the table.
|Sl No||Project||The Project Teams' understanding of the goals and objectives of the host company in alignment with the project||Presence of a Relationship management (host and partner) program and its effectiveness||Senior executive (Host company) support for the offshore outsourcing initiative||Sound and flexible contract||Total Dynamic Score for the project|
Table 2: Dynamic factors of the project
The results of the study established by the economic indicators of the projects are given below. The project functional size is given as four different classification, classified as per function point count, as project size greater than 500, less than 500 and greater than 250, less than 250 greater than 100 and small projects of functional size less than 100.
|Sl No||Project||Size in FP (IFPUG)||Project Delivery Rate in Hrs/FP||Project Duration in actual months||Original estimate of project duration in months||Percentage variation from original estimate|
|2||B||< 250 > 100||22||5||3.33||34|
|4||D||< 500 > 250||18||6||5.0||16|
|6||F||< 250 > 100||22||3.3||1.9||42|
|8 Economic Indicators of the Projects Studied||H||< 100||18||1||.8||20|
The table below indicates the relationship between the dynamic factor and the economic indicators of the projects under study.
|Project||Dynamic Score||PDR||Project Duration||Percentage Variation from the Original Estimate|
- There is a relationship between the dynamic score (DS) of the studied project and the PDR and the percentage variation in duration of the project from the estimate.
- Higher the DS, the better the productivity (lower PDR)
- Higher the DS, the lesser the variations from the estimates of project duration.
- For a score of 16 or more, it is noted that the variation could be kept below 20 %
- One-way to over come a low DS is to move to a mostly offsite engagement, that will improve PDR and reduce the variation in project duration.
We suggest utilizing the framework of a dynamic score as a review tool for the project managers during project evaluations.
Offshore outsourcing has become a feasible option for organizations to retain and gain a competitive edge in the global marketplace. Offshore outsourcing was a concept that started in 1990, and has now become a trend as it harnesses the power of information technology from far away locations to bring in economies of scale and cost competitive operations for implementing and supporting client's mission-critical projects. In today's high-tech world, the terms ‘onsite’, ‘offsite’, and ‘offshore’ merely refer to the physical locations.
Thanks to the advancement in technology, the terrestrial and satellite dedicated links, video-conferencing, online chats, e-mails, and telephones, a dual/multi-shore delivery system works as a single virtual unit, regardless of where people are actually located. So, off shoring projects alone is not the question of viability, it is the factors and risks involved in completing a project successfully. Proper understanding of the relationship between dynamics and economics would give the project manager a tool to assess the risks in their projects. A project manager must comprehend the full scope of dynamics and economic indicators to see a project to fruition.
Ettenberg, E. (2002) The next economy: will you know where your customers are? McGraw-Hill.
Pressman, R. S. (1997) Software engineering a practitioner's approach. McGraw-Hill
McCarthy, J. (1995) Dynamics of software development. Microsoft Press.
Yourdon, E. (2002) Byte Wars, The impact of September 11 on Information Technology. Prentice Hall.
Amor, D. (2000) The E-business (R)evolution, living and working in an interconnected world. Hewlett-Packard Professional Books.
International Software Benchmarking Standards Group (ISBSG) Item, The Software Metrics Compendium. from http://www.totalmetrics.com/tmdl/Compendium.pdf
International Software Benchmarking Standards Group (ISBSG) Practical Project Estimation. ISBSG.