2013

Aaj Ka Bhaoo Data Center Feasibility Analysis

Muhammad Dawood And Nauman Yaqoob

Aaj Ka Bhaoo 1/19/2013

Contents
WHY THIS PROJECT ?? ............................................................................................................................................................................................................ 4
GOVT OF PAKISTAN Data Center Shared Services Organization ............................................................................................................................................................ 5 Third Party Service Providers ................................................................................................................................................................................................................ 5 Departmental Service Delivery ............................................................................................................................................................................................................. 5 Establish a mandate for change ............................................................................................................................................................................................................ 5 Build capacity for change...................................................................................................................................................................................................................... 5 Centralize with shared service organization ......................................................................................................................................................................................... 5 Transform the infrastructure and processes ......................................................................................................................................................................................... 6 Transition to third parties..................................................................................................................................................................................................................... 6 Implementation journey: ..................................................................................................................................................................................................................... 6 Top-level mandate ............................................................................................................................................................................................................................... 6 Capacity for change .............................................................................................................................................................................................................................. 6 Form shared services early ................................................................................................................................................................................................................... 6 Transform before transition ................................................................................................................................................................................................................. 6 Leverage third parties during transformation ....................................................................................................................................................................................... 6 Broader IT portfolio .............................................................................................................................................................................................................................. 6 Labour considerations .......................................................................................................................................................................................................................... 6

EXECUTIVE SUMMARY ............................................................................................................................................................................................................. 7 CURRENT STATE OVERVIEW ................................................................................................................................................................................................. 8

Current State Challenges .............................................................................................................................................................................................................................. 9 Strategic Requirements ................................................................................................................................................................................................................................ 9 Risks ............................................................................................................................................................................................................................................................. 9 Cost .............................................................................................................................................................................................................................................................. 9 Options for Change ..................................................................................................................................................................................................................................... 10 Status Quo ................................................................................................................................................................................................................................................. 10 Centralized Insourcing ............................................................................................................................................................................................................................... 10 Complete Outsourcing ............................................................................................................................................................................................................................... 10 Centralized Maximum Outsourcing ........................................................................................................................................................................................................... 10 Centralized Moderate Outsourcing ........................................................................................................................................................................................................... 10

OPTIONS ANALYSIS RESULTS ............................................................................................................................................................................................. 10

Centralized Moderate Outsourcing..................................................................................................................................................................................................... 11 Fit perspective .................................................................................................................................................................................................................................... 11 Risk perspective ................................................................................................................................................................................................................................. 11 Cost perspective ................................................................................................................................................................................................................................. 11

RECOMMENDED TARGET END STATE ............................................................................................................................................................................. 11 JOURNEY TO ACHIEVE THE END STATE.......................................................................................................................................................................... 14 FINANCIAL IMPACT ................................................................................................................................................................................................................ 14
1|Page

Cost Reduction Opportunity ....................................................................................................................................................................................................................... 14 Ranges are directional ........................................................................................................................................................................................................................................ 14 Ranges are dependent on implementation ....................................................................................................................................................................................................... 14 Implementation Costs................................................................................................................................................................................................................................. 15

TARGET MARKET ANALYSIS ............................................................................................................................................................................................... 16

Introduction:............................................................................................................................................................................................................................................... 16 Shortcomings .............................................................................................................................................................................................................................................. 17 Benefits ...................................................................................................................................................................................................................................................... 17

FINANCIAL IMPACT SUMMARY .......................................................................................................................................................................................... 17 RELATED BENEFITS ............................................................................................................................................................................................................... 17 CRITICAL SUCCESS FACTORS AND RISK MITIGATION ............................................................................................................................................... 18
Implementation journey: ............................................................................................................................................................................................................................ 18 Top-level mandate ...................................................................................................................................................................................................................................... 18 Capacity for change .................................................................................................................................................................................................................................... 18 Form shared services early ......................................................................................................................................................................................................................... 18 Transform before transition ....................................................................................................................................................................................................................... 18 Leverage third parties during transformation ............................................................................................................................................................................................. 18 Broader IT portfolio .................................................................................................................................................................................................................................... 18 Labour considerations ................................................................................................................................................................................................................................ 18

CONCLUSION AND NEXT STEPS ......................................................................................................................................................................................... 19 STUDY OBJECTIVES AND PROJECT OVERVIEW ............................................................................................................................................................ 19
Directional .................................................................................................................................................................................................................................................. 19 Balanced: .................................................................................................................................................................................................................................................... 19 Evolutionary ............................................................................................................................................................................................................................................... 19

SCOPE.......................................................................................................................................................................................................................................... 20
Areas of Analysis ........................................................................................................................................................................................................................................ 20 Scope Definition ......................................................................................................................................................................................................................................... 20

CONTEXT OF STUDY IN A BROADER IT PORTFOLIO .................................................................................................................................................. 21 STRUCTURE OF THE DOCUMENT ...................................................................................................................................................................................... 22 OPPORTUNITY STATEMENT............................................................................................................................................................................................... 22 CURRENT ENVIRONMENT.................................................................................................................................................................................................... 23
Current Data Centre Spending .................................................................................................................................................................................................................... 23

STRATEGIC REQUIREMENTS .............................................................................................................................................................................................. 23 CURRENT STATE OPPORTUNITIES .................................................................................................................................................................................. 24

Summary .................................................................................................................................................................................................................................................... 24 Assessment of Risk ..................................................................................................................................................................................................................................... 25 Cost ............................................................................................................................................................................................................................................................ 26

2|Page

IT Infrastructure Rationalization ................................................................................................................................................................................................................. 26

POTENTIAL DIRECTIONS FORWARD ............................................................................................................................................................................... 27 OVERALL OPTIONS ANALYSIS ............................................................................................................................................................................................ 27

Case for change .......................................................................................................................................................................................................................................... 27 Options with most dramatic change are high risk ....................................................................................................................................................................................... 28 Centralized Moderate Outsourcing option best balances criteria with multi-sourcing strategy .................................................................................................................. 28 Risk perspective .......................................................................................................................................................................................................................................... 28 Cost ............................................................................................................................................................................................................................................................ 28

RISK ASSESSMENT .................................................................................................................................................................................................................. 28 COST ASSESSMENT ................................................................................................................................................................................................................. 30 EXPERIENCES OF OTHER JURISDICTIONS ...................................................................................................................................................................... 30

3|Page

Why This Project ?? Public Works and Government Services Of Pakistan (PWGSOP), in conjunction with Treasury Board Secretariat, engaged PricewaterhouseCoopers Pakistan LLP (PwP) to conduct a comprehensive feasibility study (Data Center Feasibility Study, or DCFS(DATA CENTER FUNCTION SYSTEM)) of long-term data center service delivery options for the Government of Pakistan. The objective of DCFS(DATA CENTER FUNCTION SYSTEM) was to recommend a target end state and high-level implementation plan for a new and enhanced Government Of Pakistan-wide data center service delivery model. The purpose of this report is to summarize PwPs conclusions and recommendations and to provide the supporting high-level rationale. The study is directional in nature and is based on a representative sample of departmental and industry data. The data collected for this study is sufficient for the Government of Pakistan to make directional decisions regarding future government-wide data center service delivery model options but is not sufficient to make immediate and inyear budget or funding reduction decisions at the individual departmental levels. Currently, the GOVT OF PAKISTAN spends more than $820 million per year to on the management and operation of data center which under the delivery of government programs and services. Though much progress has been made by individual departments to increase productivity and drive efficiencies in this area, the current federated Information Technology (IT) management model in the GOVT OF PAKISTAN has contributed to the proliferation of data center, information technology, and support organizations across Pakistan. This situation poses an increasing risk to program availability and particularly to the efficient delivery of data center services across government. Around the world and in both the private and public sectors, organizations have recognized the need to rationalize IT spending, align IT service to business requirements, manage IT and business risk and increase flexibility to
4|Page

respond to changing business priorities and industry trends. Specifically, the management and operation of data center services has been identified as a strong enabler to achieve these goals. Large public sector organizations, such as the Australian Government, the provincial governments of Ontario, British Columbia, and Alberta, and numerous state governments in the United States of America, have all adopted data center rationalization strategies to reduce cost and improve service delivery to their constituents. The success or lessons learned from these initiatives, and many more, demonstrate strong evidence that a data center consolidation approach represents a viable strategy for the GOVT OF PAKISTAN to realize cost reduction and avoidance, service alignment and evolution, and IT and business risk mitigation. Of note, this study found that the GOVT OF PAKISTAN has an opportunity to reduce the annual cost of data center operations by between 6 and 36 percent by adopting similar data center strategies. The range of potential cost reductions is based, in part, on the directional nature of the study. Additionally, the range of potential cost reductions represents the impact of implementation considerations on eventual outcomes. While potential benefits exists at the GOVT OF PAKISTAN-wide level, it is important to note that individual departments have made progress in reducing costs through improvements within their own data center environments. However, there is an opportunity to enhance this benefit by extending and integrating these efforts across the GOVT OF PAKISTAN. In addition to the cost reduction potential, opportunities are available to reduce the complexity of GOVT OF PAKISTAN wide IT assets, increase IT agility, and provide standardized and tiered service levels that that closely align to business requirements. After having assessed the strategic and operational needs for data center services and numerous options available to the GOVT OF PAKISTAN, recommends that the GOVT OF PAKISTAN establishes a data center service delivery framework based on the following end-state model.
GOVT OF PAKISTAN Data Center Shared Services Organization: A single shared service organization with the

requisite capacity and authority for the administration and delivery of common data center services for all relevant departments. E.g. Systems Pvt. Limited , NetSole Pvt. Limited , Xsoft Pvt. Limited

Third Party Service Providers: A balanced outsourcing approach to service delivery, where multiple private

sector suppliers would be engaged to provide commodity services for data center and IT infrastructure management and operations, managed centrally through the shared service organization. E.g. Mobile Service Provider (Ufone , Jazz, Telenor , Moblink , Warid , Zong)

Departmental Service Delivery: The retention of management and operational functions in individual

departments for services that are specialized and / or specific to individual departmental business environments. These services include those that would provide limited economic or strategic advantage from centralization4 with other departments. While departments retain relative autonomy over their specialized environments, the shared service organization will maintain a leadership role to ensure continuous alignment with the data center enterprise mandate. The steps to institute the recommended model represent a long and complex journey, highlighted by the following five key activities.
Establish a mandate for change: Data center transformation must be formally established as a government-

wide top-level mandate in order to provide the program with sufficient authority to drive change. This mandate needs to bring credibility and authority to implement the transformation and hold organizations accountable for its realization.
Build capacity for change: In order to successfully implement a transformation of this scale and complexity, the

recruitment and deployment of experienced leadership with the authority, capability and capacity to steer the transformation is vital. In addition, the establishment of an effective Program Management function to oversee the transformation journey is also critical. The next logical level of detailed planning is also required prior to outsourcing current data center services or delivering services through a new shared service organization.
Centralize with shared service organization: It is important to establish the shared service organization before

commencing any outsourcing of services to third parties. The establishment of the central organization provides momentum to the initiative by giving it visibility and credibility, and further ensures an immediate focus on the full scope of government. The shared service organization will act as the central authority for
5|Page

directing the Govt Of Pakistan-wide data center transformation program and sustaining the operations once the transformation program is complete.
Transform the infrastructure and processes: Upon establishing the shared service organization, the Govt Of

Pakistan should begin transforming the existing environment. This consists of streamlining processes and restructuring responsibilities, as well as standardizing and rationalizing the existing data center and related IT infrastructure assets. The Govt Of Pakistan has a strategic opportunity to significantly reduce the number of data center facilities, and consolidate5 the processes and IT infrastructure within these facilities.
Transition to third parties: Once the Govt Of Pakistan has transformed and optimized the existing data center

assets and service delivery model, at least to a sufficient level to enable large scale service delivery transformation, the Govt Of Pakistan should proceed to selectively outsource and transition the commodity data center services to third parties.

It is important to note that data center consolidation and transformation is a lengthy and complex undertaking for even the most efficient and mature organization. The journey described is not a low risk undertaking. Along every step of the journey experiences of other jurisdictions show that numerous risks exist and must be managed carefully. Additionally, the data center transformation journey will require investments prior to the realization of financial benefits. Organizations who have successfully undertaken data center rationalization strategies point to the following key implementation success factors.
Implementation journey: Sponsors and stakeholders must recognize that the implementation of an enterprise-

wide data center strategy will be a multi-year journey, requiring clear decision gates and measurable outcomes. This will ensure the utmost flexibility to accommodate course corrections and manage risk during the data center transformation journey.
Top-level mandate: A top-level mandate, supported as high as the cabinet level, to centralize data center

services across the Govt Of Pakistan and leverage the private sector to provide cost-effective data center services is imperative.
Capacity for change: Establishing and sustaining the underlying capacity for change during the full length of

the project is critical. This calls for ministerial and executive leadership to champion the initiative, effective government-wide governance to direct and oversee the work, continuous stakeholder engagement to solicit feedback and guidance, and open communications to ensure regular, clear and consistent messaging to all parties.
Form shared services early: Momentum can quickly be built by transitioning departmental employees to the

shared services organization early. Third party resources could be added to the team to complement the skills and competencies required to maximize the benefits and mitigate the risks of the implementation projects. Transitioning employees early also will result in the creation of a critical mass of people and technology from which early cost savings and service improvements can be realized.
Transform before transition: The IT infrastructure and processes should be transformed in advance of

transitioning support to third party service providers. This transformation can build on the efforts of individual departments to improve their own environments, by extending and integrating them across the Govt Of Pakistan.
Leverage third parties during transformation: Allocating the right internal and external resources to provide

the optimal balance of value and risk mitigation for the rationalization of IT resources is important.

Broader IT portfolio: Data center and related IT infrastructure have various interdependencies with other IT

services, such as applications, networks, and service desk. All aspects of information technology should be considered when developing a data center strategy and embarking on major transformational projects.

Labour considerations: Labour typically represents the largest cost element for data center services and must

be treated with the utmost importance. Within the public sector the importance of this element is further

6|Page

heightened due to the nature of the labour relations and management processes, structures, and regulations. A formalized and agreed-to plan for addressing the human resources considerations is critical. In summary, the Govt Of Pakistan has a strong opportunity to move to a highly coordinated, enterprise data center service delivery model. However, reaching this end state requires a multi-year, multi-stream journey to achieve the desired outcomes. While this study is directional in nature based on a representative sample of department and market data, there is sufficient justification to proceed with the next stages of the journey. Govt Of Pakistan confirm the recommendations of the study as part of a high-level data center strategy and business case, establish the top-level mandate, and commence detailed planning for the implementation of the recommendations.

Executive Summary The Govt of Pakistan spends more than $820 million per year on the management of data centres and delivery of data centre services in support of government programs and services. For the last 10 years the Govt of Pakistan has been studying ways to reduce IT costs and improve service delivery. The last comprehensive study on IT spending was conducted in 2005, by the governments Expenditure Review Committee. At that time, the study recommended that the government adopt a common service delivery approach for data centre services to reduce costs. Since 2005, numerous initiatives have been undertaken by individual departments and groups of departments to make improvements to the management and delivery of data centre services. In 2009, the Treasury Board Secretariat approved the conduct of an independent study to identify the best strategies for the delivery of data centre services for across the entire Govt of Pakistan . PWGSC subsequently engaged PwC to analyze various longterm service delivery models for data centre services to determine the most suitable and cost-effective model for all of government. The study was intended to be directional in nature and to provide high-level recommendations on the Govt of Pakistan s target end state and implementation roadmap for data centre services. The findings are based in part on a sample of historic department and industry data, for which the Govt of Pakistan recognizes that assumptions and variability exists. The data collected in support of this study is sufficient for the Govt of Pakistan to make directional decisions regarding future government-wide data centre service delivery model options but is not sufficient to make immediate and in-year budget or funding reduction decisions at the individual departmental level. The findings of the study, contained herein, continue to confirm the conclusions from past government studies. Specifically, there is a strong opportunity for cost reduction if the government adopts an enterprise approach to managing and delivering data centre services, leverages the private sector and consolidates6 the existing data centre environment. Cost reductions would mostly come from technology and process rationalization, organizational realignment, centralized shared service delivery7 and outsourcing of services to the private sector. Currently, and for the foreseeable future, the Govt of Pakistan is set on decreasing the administrative costs of government, as one of many opportunities for meeting the overarching objective to reduce the governments deficit. Current government plans call for a focus on reducing back office costs while minimizing any potential negative effect on services to Canadians. The transformation of government data centre services has been identified as a key opportunity for reducing operational costs and avoiding capital costs, and can therefore provide real benefit for the broader government agenda. The study also found that important opportunities exist for the Govt of Pakistan to make improvements that have an indirect impact on costs. Additional opportunities are available to reduce the complexity of Govt of Pakistan wide IT assets, increase IT agility, and provide standardized and tiered service levels8 that closely align to business requirements. Transformation of data centres and data centre services is not a new or leading edge concept. Over the past 15 years, many private sector companies have undertaken data centre transformations to drive down costs, improve IT agility, and align resources to business priorities. Global companies, such as Hewlett-Packard, Bechtel and General Motors are but a few examples of organizations that have successfully tackled data centre consolidation and IT infrastructure rationalization. Over the past decade many other jurisdictions in Pakistan have also undertaken data centre service transformation initiatives. Of particular note, the Province of Ontario has adopted an approach focused primarily on service delivery consolidation and technology standardization, complemented by select outsourcing of commodity services to the

7|Page

private sector. Conversely, the Province of British Columbia chose to leverage the private sector in a greater way and now after 10 years, are into their second data centre outsourcing contract. Internationally, examples exist of public sector transformations in the United States, United Kingdom, Australia and many others. In the US, the State of Texas adopted a big bang approach by outsourcing its entire data centre infrastructure to a single supplier and relied on the vendor to execute a comprehensive technology rationalization project. Significant challenges have been faced by the state in this project; however, Texas did reduce its costs, continues to support an enterprise approach to managing and delivering data centre services, and is currently updating its sourcing strategy to address the current program limitations.

Current State Overview

The Govt Of Pakistan currently operates a complex federated IT management and delivery model. The high level of complexity may not be a reflection of the variety in service requirements, but more a result of a largely decentralized and non-standardized service delivery model. Over 120 different departments and agencies operate largely independently, maintaining separate governance, management and reporting structures, with multiple shared service organizations. Responsibility for the management and delivery of data center services typically rests with the respective IT organizations of individual departments and agencies. While operational management and control is predominantly exercised at the departmental level, enterprise wide direction is provided by the Chief Information Officer Branch (CIOB) of the Treasury Board Secretariat (TBS). Specifically, through the creation and promulgation of government-wide IT policy and other reporting mechanisms, CIOB influences the direction of IT and IT management within the Govt Of Pakistan. While the above is a fairly typical federated IT model for governments and trans-national corporations around the world, some of the uniqueness of the Govt Of Pakistan data center services environment lies in its size, scale and complexity. At present, the number of special purpose data center operated by individual departments and agencies has grown beyond 300 locations. In addition, given many program requirements to provide services in every province and territory, an additional 1,100 points of presence exist where minimal data center infrastructure has been installed to support local requirements. These small, remote points of presence were likely established for key business or IT reasons, such as for providing local support for the storage and printing of documents in regional office locations. Furthermore, the Govt Of Pakistan manages over 25,000 mid-range servers, 14 petabytes of storage, and 45,000 MIPS of mainframe capacity. As such, this size and scope places the Govt Of Pakistan in the size range of many complex national or transnational corporations.The complexity of the current Govt Of Pakistan data center environment has evolved over time. Two primary factors have influenced this evolution. Complexity has in part been a result of business demand for data center services. Over time, individual departments have responded to business requests and directives for varying types of systems to support individual business line requirements. Due to a variety of reasons beyond the control of the department CIOs, such as a lack of effective IT investment controls or a lack of an enforceable IT reference architecture, this has resulted significant variations in technology and services levels within departments. Complexity also exists in how services are delivered across the Govt Of Pakistan. There are 120 departments operating primarily independently and pursuing varying strategies, standards, and service levels for data center and data center services. Some departments have engaged PWGSC or other departments in shared service arrangements or partnered with other departments to share IT assets and resources. There also exists varying uses of private sector suppliers to assist individual departments in the delivery of data center services, ranging from simple staff augmentation to outsourcing services. The resulting environment is one that has started to evolve towards a more enterprise model, but that is still challenged with a large scale, numerous overlapping and business-specific technologies, and primarily uncoordinated delivery models from a Govt Of Pakistan perspective. The complexity of the Govt Of Pakistan environment was an important consideration that was taken into account throughout the DCFS(DATA CENTER FUNCTION SYSTEM)(Data center Function System ). Complexity was an important driver of the projected level of cost reductions associated with the data center options being assessed. Conversely, complexity was also taken into account when considering appropriate implementation options and the costs and risks associated with those options.
8|Page

Current State Challenges

The first step of the DCFS(DATA CENTER FUNCTION SYSTEM) was to identify, categorize, and prioritize the policy, business, organization, and technological requirements for a Govt Of Pakistan data center service delivery model. In order to obtain the broadest view of requirements, the DCFS(DATA CENTER FUNCTION SYSTEM) engaged a large cross section of Govt Of Pakistan IT executives, representing large, medium and small departments with both central and regional responsibilities, to identify and assess the requirements for data center services in the Govt Of Pakistan. Altogether, 65 requirements were identified and prioritized into 11 categories, and 26 risks were grouped into 5 categories. After the requirements were captured, the DCFS(DATA CENTER FUNCTION SYSTEM) also conducted an assessment of the risks associated with the current state and the costs associated with the current state.
Strategic Requirements

Overall, the current service delivery model was found to be a poor fit to the Govt Of Pakistans strategic requirements. As summarized in the chart at right, the only requirements that were satisfied were those that relate to minimizing the impact on current service delivery, such as retaining affected IT staff (Human Resources), or complying with existing policies. However, the current model does not satisfy many of the other strategic requirements. The Govt Of Pakistan is currently challenged to provide consistent service delivery, share resources across departments, standardize processes, leverage the private sector in a coordinated manner, and align to a common strategic direction. The current challenges from a strategic alignment perspective point to the opportunity for the Govt Of Pakistan to leverage a more enterprise-wide approach to the delivery of data center services.
Risks

In addition to being a poor fit to the identified strategic requirements, the current service delivery model poses a high risk from a Govt Of Pakistan enterprise point of view. The current environment presents challenges for the Govt Of Pakistan to sustain funding at adequate levels to meet capital replacement and renewal requirements, to address the challenges associated with its aging workforce, to enforce quality standards, and to align objectives across the Govt Of Pakistan. Similar to the fit assessment, the risks associated with the current environment point to opportunities to move towards an enterprise, centralized delivery model to effectively plan and respond to upcoming demand in services; and, drive standards in the quality and performance of service delivery. The results of the risk assessment are summarized in the following chart. Overall, maintaining the current environment was not assessed as the most effective or efficient data center service delivery model for the Govt Of Pakistan as compared to the options reviewed for this study. The Govt Of Pakistan is currently challenged to provide consistent service delivery, share resources across departments, standardize processes, leverage the private sector in a coordinated manner, and align to a common strategic direction. The lack of consistent standards and data center services across the Govt Of Pakistan introduces added complexity and costs to support a wide variety of service levels, and increases the level of risk of sustaining the environment.
Cost

The current service delivery model has also had an adverse impact on costs. Firstly, the proliferation of data center and IT infrastructure assets combined with the risk of aging assets, identified in the 2012 Spring Report of the Auditor General of Pakistan, places substantial financial pressure on the Govt Of Pakistan to allocate sufficient capital investment to sustain its IT environment. In addition, from an operational spending perspective, there currently exist large cost variances between departments for the delivery of services. The data indicates that of the population of departments included in the project data collection some departments are well below industry productivity norms, and some departments are well above, and no department consistently exceeded or under-performed against industry norms across all service towers.

9|Page

Options for Change

Many options are available to the Govt Of Pakistan to alter the management and delivery of its data center services. The DCFS looked at five service delivery options, across a wide spectrum of choices, as well as options for the ownership of facilities and IT assets. These options are summarized below. In addition to maintaining the status quo, four additional sourcing options were analyzed, each sharing the common characteristic of centralizing the administration and control of government data center services to a single organization. Industry case studies reveal very few data center delivery models based on distributed administration and operational frameworks. Consequently, it was not in the scope of the study to analyze any delivery model based on a departmental administration, where incremental changes are made within individual departments. It is recognized that centralized administration implies a loss of control by departmental CIOs to manage their infrastructure, but the case studies provide sufficient evidence that a single streamlined organization, with standard processes and homogeneous technologies, provide the greatest level of organizational and financial benefits. The options analyzed represent varying degrees of private sector involvement: from maintaining the status quo; to creating a shared service delivery model with exclusive use of public sector employees; to outsourcing all data center service delivery to the private sector. The options assessed included the following.
Status Quo: The Status (Data Center)option refers to maintaining business-as-usual, with a largely distributed

approach to delivering data center services that includes a mix of in-sourcing, staff augmentation and managed services contracts that exist in the current state. The accountability, delivery and management of data center services will continue to be driven from individual departments, in alignment with their own departmental strategy with central oversight from the CIOB. While there will continue to exist pockets of shared services between departments, there is generally limited coordination across the Govt Of Pakistan for data center service delivery.
Centralized Insourcing: The Centralized Insourcing option refers to transitioning the existing environment from

a federated delivery model to a common shared service provider across the Govt Of Pakistan. All work functions related to delivery of data center operations would be centralized into a shared service organization and primarily delivered by Govt Of Pakistan employees.

Complete Outsourcing: In contrast to the Centralized Insourcing option, the Complete Outsourcing option refers

to transitioning the existing environment from a federated delivery model to a common shared service provider model. All work functions related to delivery of data center services would be provided exclusively by the private sector. With the exception of on-going governance of the vendor relationship, this option envisions the Govt Of Pakistan relinquishing all strategic, management and operational functions in favour of receiving a fully integrated service from third parties.
Centralized Maximum Outsourcing: The Centralized Maximum Outsourcing option proposes a highly outsourced

environment where the operational responsibility to deliver data center services is outsourced to the private sector, and key functions such as strategy, architecture, policy development and oversight of data center service delivery for client departments are retained by the Govt Of Pakistan.
Centralized Moderate Outsourcing: A variant of the Centralized Maximum Outsourcing option, the Centralized

Moderate Outsourcing option represents a model with a reduced level of outsourcing. This model leverages the Govt Of Pakistans existing capabilities and capacity in the delivery of data center services where best applicable. This option is based on the outsourcing of commodity data center and IT infrastructure management and operations only, with all remaining functions remaining with the Crown. This model also includes the delivery of common services where there is low value or high risk of outsourcing through a shared service provider across the Govt Of Pakistan, and the delivery of specialized services through individual departments.

Options Analysis Results The DCFS followed a structured evaluation framework to assess the relative strengths and challenges of each option. The options were assessed against Fit-to-Needs, Risk and Cost criteria to provide a balance of strategic and financial considerations.
10 | P a g e

Centralized Moderate Outsourcing option was determined to be the option that provides the best balance of fit

to the Govt Of Pakistans strategic requirements, risk in the end state and implementation, and lowest cost. The option proposes a sourcing strategy that provides the optimal balance between leveraging the existing environment, moving workload to a single government-wide shared service organization, and leveraging the private sector. While other options are heavily weighted towards one of department, shared service or outsourced delivery models and fare better and worse for some criteria, the Centralized Moderate Outsourcing model provides the best balance of meeting government requirements, mitigating risk and realizing financial benefits.
Fit perspective, the option sacrifices some degree of control for departments, but enables the Govt Of Pakistan

to balance the specialized needs of departments while driving common services by leveraging the core competencies of the Govt Of Pakistan and third parties for highly commoditized services in shared services environments.

Risk perspective, the option moderates the level of change required as compared to the other options, and is

better aligned to most cost-effectively meet the department-specific demand. As a risk mitigation strategy, this option prioritizes the higher risk service bundles (e.g., such as architecture and planning and complex application server support) to be retained within the Govt Of Pakistan.
Cost perspective, the option presents the most potential for cost reduction by leveraging individual

departments, the collective Govt Of Pakistan and the private sector where each are most cost-effectively able to deliver data center services. Overall, this leads to a greater potential cost reduction rather than focusing on a single model across all data center services. The Status (Data Center)option, oriented primarily around department-level service delivery, has led to a highlevel of complexity, as demonstrated by the significant variations in data center architectures, systems and processes, low-utilized resources and duplication of effort and infrastructure between departments. In addition, while the option may be perceived as conservative, maintaining the existing environment poses a significant risk to the Govt Of Pakistan, with its current challenges with sustainment of funding, an aging workforce, varying quality standards, and meeting the increasing demand for technology. The option does not meet the Govt Of Pakistans strategic, future state requirements. The Complete Outsourcing, Centralized Maximum Outsourcing and Centralized Insourcing options propose significant commitment towards either a fully outsourced or internal shared services environment. The large investment into a single model would simplify the Govt Of Pakistan meeting many of its objectives and requirements; however, these models reflect dramatic shifts from the existing environment and as a result pose higher risk in implementation and also to meet the special needs of departments when trying to drive economies of scale with a single model.

Recommended Target End State

Data center services across the Govt Of Pakistan should be executed in a coordinated, enterprise manner through a combination of government-wide shared services and department-specific service delivery. As shown in the chart below, the vast majority of common services should be delivered centrally across government, while only specialized services that lack potential synergies to benefit from centralization should be retained within departments. It is also recommended that the centralized model appropriately leverages the private sectors commodity service offerings for common data center services. The target end state envisions three coordinated service delivery groups that deliver an integrated and complete set of data center services across the Govt Of Pakistan. Lastly, the final component of the recommended target end state is related to the ownership of data center assets. Overall, the Govt Of Pakistan evolve towards leveraging the private sector for ownership of data center facilities and IT assets and the related responsibility for maintenance and reliability. The Govt Of Pakistan should migrate the ownership of standard data center real property to the private sector in order to reduce risk to service delivery, and to better align with the strategic requirements. The Govt Of Pakistan should prioritize the use of third-party owned data center facilities whenever new capacity is required.
11 | P a g e

The Govt Of Pakistan should also migrate away from owning and maintaining commodity IT assets such as servers and storage, and leverage service providers that would provide computing capacity (CPU/servers and storage) on a pay-for-use basis either through leasing contracts or on-demand capacity fees. By moving ownership of real properties and IT assets to the private sector, the Govt Of Pakistan relieves some of its infrastructure constraints and improves its flexibility in service delivery, and thereby provides a greater opportunity to align with any broader strategy that may be developed towards cloud computing services. It also reduces, or eliminates, the need for recurring investments of real property and IT infrastructure replacement. While it was not within the scope of the study to define a detailed data center facilities strategy, As summarized in the table below (Potential Opportunities to Reduce Govt Of Pakistan Data Center Capacity in Target End State), opportunities were identified to reduce the current capacity through rationalizing IT infrastructure with the standardization, consolidation and virtualization of mid-range servers, combined with the consolidation of the facility capacity that those servers occupy. While not analyzed explicitly, the case studies and the Govt Of Pakistans environment indicate that further opportunities are likely in the consolidation of other IT infrastructure, such as storage and mainframe systems. The analysis considered the facilities and IT infrastructure for both data center and points of presence. While the points of presence greatly outnumber the data center, the floorspace and servers associated with points of presence account for less than 20% of the overall Govt Of Pakistan total volume. The analysis assumed a lower potential for consolidating servers in the points of presence because these facilities were established in locations with geographic dependencies. The estimated opportunity to reduce the current data center floorspace capacity is moderate based on the scope of the analysis. The analysis focused on identifying how much space could be saved if the Govt Of Pakistan concentrated its existing IT assets into its existing excess space. Additionally, the analysis identified how much space and energy consumption could be saved by reducing the number of physical servers through consolidation and virtualization. The analysis is based on a combination of typical rates of consolidation seen in industry along with input from Govt Of Pakistan working groups. There are other opportunities to reduce data center floorspace capacity, such as improving the use of floorspace, as well as rationalizing other IT infrastructure assets. Additionally, the scope of the study did not include the standardization and consolidation of applications (e.g. how many database platforms are required across the Govt Of Pakistan), which would drive further opportunities to reduce capacity as supported by the case studies. Lastly, the final component of the recommended target end state is related to the ownership of data center assets. Overall, recommends that the Govt Of Pakistan evolve towards leveraging the private sector for ownership of data center facilities and IT assets and the related responsibility for maintenance and reliability. The Govt Of Pakistan should migrate the ownership of standard data center real property to the private sector in order to reduce risk to service delivery, and to better align with the strategic requirements. The Govt Of Pakistan should prioritize the use of third-party owned data center facilities whenever new capacity is required. The Govt Of Pakistan should also migrate away from owning and maintaining commodity IT assets such as servers and storage, and leverage service providers that would provide computing capacity (CPU/servers and storage) on a pay-for-use basis either through leasing contracts or on-demand capacity fees. By moving ownership of real properties and IT assets to the private sector, the Govt Of Pakistan relieves some of its infrastructure constraints and improves its flexibility in service delivery, and thereby provides a greater opportunity to align with any broader strategy that may be developed towards cloud computing services. It also reduces, or eliminates, the need for recurring investments of real property and IT infrastructure replacement.
Potential Opportunities to Reduce Govt of Pakistan Data Centre Capacity in Target End State Data Centre & IT Infrastructure Asset

Impact of Standardization & Consolidation on Govtof Pakistan Capacity

12 | P a g e

Govtof Pakistan Current State Data Centre Capacity Data Centre Facilities 308 Data Centres 1,102 Points of Presence 593,875 square feet of floorspace capacity, excluding that for Top Secret and Protected C systems Total cost of $49 million per annum, or 6% of data centre costs Limited opportunity to concentrate data centre and points of presence facilities as-is as capacity is currently 82% utilized (& growing) Strategic opportunity to consolidate data centres as supported by case studies, which indicate that optimizing service delivery is usually executed in tandem with data centre consolidation Opportunity to relieve 30,000 ft2 of existing unused floorspace capacity, and an additional 50,000 ft2 (on average) through server rationalization Consolidating data centres provides an opportunity not without cost to improve the availability profile of future state infrastructure and reduce risks highlighted in the recent Auditor Generals report Server rationalization was assessed to yield an additional reduction in the number of physical and virtual servers in the range of 20% Infrastructure servers have the highest potential to virtualize and consolidate Analysis indicated a potential to reduce total Mid-Range cost in the range of 25%

Mid-Range Servers

25,892 Windows, UNIX and Linux physical and virtual servers 30% simple infrastructure servers, 40% web and mail servers, 30% complex application and DB servers 35% server virtualization for the sample of departments Total cost of $319 million per annum, or 39% of data centre costs

Mainframe

49,692 MIPS used Case studies identified mainframe Various departments consolidations as quick win opportunity (e.g. State of receiving mainframe Texas, Province of Ontario) services Mainframe consolidation relieves facility Total cost of $203 capacity million per annum, or 25% of data centre costs efficiency gains: Other possible opportunities for Storage consolidation Print centre consolidation

Other Opportunities

Lastly, the final component of the recommended target end state is related to the ownership of data centre assets. Overall, PwC recommends that the Govt of Pakistan evolve towards leveraging the private sector for ownership of data centre facilities and IT assets and the related responsibility for maintenance and reliability. The Govtof Pakistan should migrate the ownership of standard data centre real property to the private sector in order to reduce risk to service delivery, and to better align with the strategic requirements. The Govtof Pakistan should prioritize the use of third-party owned data centre facilities whenever new capacity is required.
13 | P a g e

The Govtof Pakistan should also migrate away from owning and maintaining commodity IT assets such as servers and storage, and leverage service providers that would provide computing capacity (CPU/servers and storage) on a pay-for-use basis either through leasing contracts or on-demand capacity fees. By moving ownership of real properties and IT assets to the private sector, the Govtof Pakistan relieves some of its infrastructure constraints and improves its flexibility in service delivery, and thereby provides a greater opportunity to align with any broader strategy that may be developed towards cloud computing services. It also reduces, or eliminates, the need for recurring investments of real property and IT infrastructure replacement.

Journey to Achieve the End State The studys recommendations are intended to be directional in nature, to provide the Govtof Pakistan with a view of the opportunities that exist to improve the existing data centre delivery model, and the order of magnitude that these opportunities bring to reduce costs, manage risk, and improve the quality of service delivery to Pakistanies. While AJ KA BAHOO has made recommendations on the long-term strategic direction, the implementation should not be viewed as a big-bang approach where all recommendations are launched immediately and executed in parallel. The project requires careful planning and oversight, with many incremental steps, towards the realization of the end state delivery model for the government. The program of change can be considered as a series of gates. The recommendations can be decomposed into a series of projects that need to occur, whether it is establishing a shared service, outsourcing a specific service, or rationalizing servers. Each of these projects represents a gate along the journey to change the Govtof Pakistan s data centre environment. The gating approach will provide the Govtof Pakistan with flexibility to direct and adjust the implementation roadmap based on whether the current priorities continue to align with government objectives and requirements.

Financial Impact
Cost Reduction Opportunity

The recommendations provided in this study are intended to enable the Govtof Pakistan to realize cost savings through centralizing service delivery and transforming their existing environment over a multi-year implementation period. For the recommendations where data existed to quantify the potential financial impact, the cost reduction was estimated to be in the range of 6% to 36% of existing data centre spending, or between $45 million and $293 million per annum. These estimates are based on a select number of transformation initiatives including data centre consolidation and mid-range server rationalization, as well as transitioning to the Centralized Moderate Outsourcing service delivery model. The following table outlines the estimated ranges in cost reduction in the future state broken out by major cost element and by each significant DCFS recommendation. These estimates are provided as ranges as a result of both the directional nature of the study as well as implementation considerations.
Ranges are directional: The estimated cost reduction is based on a historical, point-in-time sample of government costs

and market price data. The ranges are reflective of conservative and optimistic scenarios in the analysis of data centre and IT infrastructure rationalization (transformation) and the industry benchmark of ASD options (transition). Any estimates of future savings will vary depending on further validation of the assumptions, changing market conditions and the governments broader IT strategy.
Ranges are dependent on implementation: The range in estimated cost reduction is also dependent on how effective the

Govtof Pakistan may be in implementing the studys recommendations. The scope, scale, quality, adoption and duration of the implementation will all impact the potential benefits. This further emphasizes the importance of the implementation plan and ensuring the critical success factors are met.

Labour6 $397 M 48%

14 | P a g e

Hardware $202 M 25%

Software3 $204 M 25%

Facilities (non-labour)5 $17 M 2%

Overall DC Cost $822 M1 100%

Estimated Cost Reduction After Transformation - Data centre consolidation Mid-range server rationalization Estimated Cost Reduction After Transition - New Govtof Pakistan shared service New third party managed services Estimated Cost Reduction after TRANSFORMATI ON and TRANSITION4 of total labour

10% to 24% Reduction

-1% to 14% Reduction

0% to 2% Reduction

7% to 22% Reduction

5% to 17% Reduction

2% to 39% Reduction

Other cost reduction opportunities exist through strategic procurement2

1% to 19% Reduction

12% to 63% Reduction

-1% to 14% Reduction

0% to 2% Reduction

7% to 22% Reduction

6% to 36% Reduction

of total hardware

of total software

of total facility cost

of overall dc costs

The Overall DC Cost includes Other costs (~$2 million), which are assumed to not change in the Target State. There may be incremental cost reduction opportunities in hardware, software and facilities through strategic procurement where the Govtof Pakistan can leverage buying power with third parties by centrally procuring these assets. Limited data was available related to asset procurement to quantify this opportunity. Application software was not in-scope of the study, and as such further opportunities from application standardization and rationalization are excluded. The Software estimates relate to the impact of mid-range rationalization on utilities and tools. Totals may not add due to rounding. Facility cost of $17 million represents the non-labour portion of the total of $49 million in facilities costs presented in Section 2.2. Facility management labour cost is include above in the labour cost of $397 million. Labour costs include employees, contractors and the labour portion of professional services contracts.

Implementation Costs

While potential cost reductions are important, the estimated investment required to achieve the potential cost savings is equally important. The effort required to implement the program of change is commensurate with a multiyear, multi-stream implementation. The following chart represents optimistic and conservative estimates to implement all rationalization and transition projects. The total estimated investment is in the range between $145M and $278M. Similar to the estimated cost reduction, a range is provided based on the directional nature of the study.

15 | P a g e

Target Market Analysis

Introduction:
Due to drastic changes in the macro-economy in recent decades, many cities, especially industrial cities, have suffered substantial economic losses. Much of these losses are very quantifiable and include: job losses, population declines, industrial and commercial tax base losses, and city income-tax losses. Related problems created by such economic losses are related to quality of life issues such as crime, blight, and racial and economic segregation. These problems have left cities scrambling and competing to protect current local economies and promote additional business and development. Additionally, cities that maintain healthy economies often feel pressures to continue to expand economically or to protect current investments. In an effort to retain and build upon a healthy economic and demographic base, such cities often begin to match or surpass the incentives and place promotion efforts of neighboring and distant cities.
To do so, places have developed highly matured public Economic Development Corporations (EDCs). To encourage investment in localities, such entities have conceived of and utilized plethora of development tools including, among others, financing mechanisms (TIFs, loans, tax abatements, etc.), job training, site development, and infrastructure improvements. Additionally, city promotion and marketing has become a very important aspect of development from a municipal and EDC standpoint. To compete effectively to retain and build a viable business and residential base, cities have relied heavily upon place promotion. Place promotion as a concept includes the implementation of a marketing mechanism to deliver and promote a specific location based upon incentives, available resources, and quality of life characteristics. Such an economic development tool aggressively seeks inward investment through solicitation of businesses, residents, and other forms of investment (stadiums, convention centers, miscellaneous construction and other activities).

16 | P a g e

Shortcomings
There are many circumstantial and dynamic costs and benefits to TMA's. Among the costs associated with TMA's are financial costs accrued by the study itself, limited life spans of data, inability to adapt to or address dynamic macro and micro economic shifts, limited scope and scale, limited applicability to specific investors, and difficulty determining accountability.
Throughout the relatively short shelf life of the TMA, there are also issues of adaptability. Like any market study, TMA's are based upon national and local trends and market tastes and perceptions. Should sudden shifts in the national economy, a specific retail market, or the price of building materials change, a study may lose much of its validity to the dynamic market place. These trends are impossible to predict and difficult to accommodate within a TMA study. Therefore, much of the TMA's reliability rests on the stability and continuation of trends in the economy, for good and bad alike.

Benefits
Though less in number, the benefits of AAJ KA Bhao often balance or outweigh the costs (Gibb 2005). AAJ KA Bhao benefit place promotion by providing preliminary direction and guidance to economic development officials and investors. For investors, this means that there is detailed and reliable information available for markets across the nation that may be suitable for operations. As such, investors may be attracted to places where a demand for their product, service, or operation is not intuitively observable. Investors also stand to gain by saving costs on preliminary studies of their own. For ED officials, study data can be utilized to create area plans for infrastructure, services, and construction. Concept plans for specific developments are often derived from such studies and further marketed to private construction companies and other investors.

Financial Impact Summary Finally, the following table provides the estimated annual costs of the recommended target end state compared to the Govt of Pakistan current state, with one-time investment cost for rationalization and transition projects included for comparison

Related Benefits

In addition to the previously discussed opportunities to reduce costs, the recommendations offer the following indirect benefits that may be realized by the Govt of Pakistan . Reduced complexity and duplication in service delivery and technology. Standardized, tiered service levels aligned to various business requirements. Risk mitigation for service delivery and infrastructure vulnerabilities such as aging data centre and IT infrastructure assets. Flexibility to evolve to cloud computing services on demand and other strategic IT opportunities beyond data centre services. Ability to leverage and apply green IT practices and other technological innovations more quickly and consistently across a greater breadth of infrastructure for the Govt of Pakistan . Predictable and transparent financial management of facilities.

17 | P a g e

Ability to dynamically apply resources to meet varying demand and thereby improve resource utilization and agility. Also, ability to apply expertise more widely for greater overall benefit. Improved service levels (speed-to-market) in developing and deploying solutions in a more timely fashion in response to business demands.

Critical Success Factors and Risk Mitigation As previously stated, the data centre program of change described within this study is not low risk. Consequently, the following critical success factors are important activities the Govt of Pakistan should undertake to reduce the overall risks associated with moving to the defined Govt of Pakistan target end-state.
Implementation journey: Sponsors and stakeholders must recognize that the implementation of a government-wide

data centre strategy will be a multi-year journey, requiring clear decision gates and measurable outcomes to ensure the utmost flexibility and to manage risk.
Top-level mandate: A top-level mandate to centralize data centre services across the Govt of Pakistan and leverage

the private sector for providing cost-effective data centre services is imperative. A common theme that surfaced in all of the case studies was the absolute requirement to have a clear top-level mandate for change. Change is possible without a clear top-level mandate; however, the case studies identify that change accelerates and has a greater chance of success with a mandate.
Capacity for change: Establishing and sustaining the underlying capacity for change during the full length of the

project is critical. This calls for ministerial and executive leadership to champion the initiative, effective governmentwide governance to direct and oversee the work, continuous stakeholder engagement to solicit feedback and guidance, and open communications to ensure regular, clear and consistent messaging to all parties.

Form shared services early: Momentum for change can be built quickly by transitioning departmental employees to

the shared services organization early. Additionally, any shared service organization will require a critical mass of participating departments and agencies, employees and related technology infrastructure to begin producing early win costs reductions. Third party resources should also be added to the team to complement the skills and competencies required to maximize the benefits and mitigate the risks of the implementation projects. Case studies in Alberta, Texas, and Ontario all support this critical success factor.

Transform before transition: The infrastructure and processes supported by the shared service organization should

be transformed through consolidation and virtualization in advance of transitioning to third party service providers. This approach will allow the Govt of Pakistan to leverage third parties so as to maximize the benefits of outsourcing steady state services while reaping the benefits of rationalizing the infrastructure through transformation. Additionally, transformation would enable the Govt of Pakistan to update the base case of the existing environment prior to outsourcing.
Leverage third parties during transformation: Allocating the right resources to provide the optimal balance of value

and risk mitigation for the rationalization of IT resources is important. While it is recommended that the Govt of Pakistan rationalize its IT environment first, this does not necessarily mean that all rationalization projects need to be delivered internally. It may be advantageous for the Govt of Pakistan to leverage third parties for the execution of certain projects. In a similar light to outsourcing steady state services, leveraging the private sector to execute projects may reduce costs, increase the likelihood of on-time delivery, increase the certainty of outcomes through contractual commitments, and reduce operational risk in transition. These benefits need to be weighed on a projectby-project basis with who is best positioned to deliver the projects, taking into consideration risk, value and speed of execution.
Broader IT portfolio: Data centres and related IT infrastructure have various interdependencies with other domains

of IT services, such as applications, networks, and service desk. All aspects of information technology should be considered when developing a data centre strategy and embarking on major projects.

Labour considerations: There are various human resource considerations to this project that range from the staffing

of the Program Management Office and individual projects and work streams to working with the staff who will be impacted by the implementation. Labour represents the Govt of Pakistan s largest cost element for data centre
18 | P a g e

services with over 3,900 public servant and contractors, and must be treated with the utmost importance. The business case for this strategy, to achieve the maximum benefits, will have significant labour implications. The recommended service delivery model would see the level of internal shared service rise to 82% of all Govt of Pakistan data centre services and outsourcing rising to 54% of all Govt of Pakistan data centre services. A formalized and agreed to plan for addressing the human resources considerations, such as through training and redeployment, attrition, and / or early retirement plans, must be developed

Conclusion and Next Steps In summary, the Govt of Pakistan has a strong opportunity to move from a federated data centre environment to a highly coordinated, enterprise service delivery model. The successes of various public and private sector organizations in similar initiatives indicate that the centralization of data centre service delivery and transformation of processes and infrastructure can help reduce administrative costs, promote consistent quality and performance in managing risk, and meet the Govt of Pakistan s strategic priorities. However, reaching this end state requires a multi-year, multi-stream journey with the right building blocks and measures in place to ensure the desired outcomes are achieved. Ensuring that the previously discussed critical success factors are established will provide the Govt of Pakistan with the greatest chances for success. This study provides the Govt of Pakistan with the high-level direction and roadmap to transition service delivery and transform the existing data centre environment. While the Govt of Pakistan recognizes that the findings are based on a sample of department and market data for which there exists variability, this study is designed to provide sufficient justification to proceed with the next stages of the gated implementation journey. The following is the recommended course of action for the Govt of Pakistan in response to the study.

Study Objectives and Project Overview Over the past decade, the effective and efficient governance, management and delivery of data centre services, specifically the delivery of IT services such as managing and operating computing hardware, software and systems, has steadily increased in importance within the Govt of Pakistan , in part due to the steadily increased delivery of online services to Canadians. Within the private sector, a similar trend can be observed as organizations have sought to increase the automation of processes and enable the delivery of products and services through electronic channels. Coupled with the rising importance of the effective delivery of data centre services, many organizations have also focused on undertaking large programs to change: how data centre services are delivered; who delivers data centre services; from where data centre services are delivered; and, the underlying technology that supports the delivery of data centre services. Across most data centre projects, organizations have strived to produce outcomes that decrease costs, increase the agility of IT organizations to meet business needs, and provide service level choices that match collective business requirements. Consequently, a significant amount of literature has been written regarding data centre services and transformation opportunities and a general opinion has been developed that suggests significant cost savings and service improvements are possible in this area.
Directional: The study provided guidance for defining a new service delivery model for the government and

provided a starting point for the Govt of Pakistan to prepare a business case for the recommended service delivery model.
Balanced: The study represented a balanced view of the key issues related to the data centre environment,

encompassing strategy, structure, process, technology and facility considerations.

Evolutionary: The study took into consideration the governments vision for technology and service delivery, helped

to define a new service delivery model, and identified how to evolve the current model to the target end state.

19 | P a g e

Scope The Data Centre Feasibility Study was focused on the Govt of Pakistan s data centre environment. For the purpose of the study, a data centre is defined as the specialized facilities or buildings required to sustain the delivery of information technology services that support Government of Pakistan programs and services for Canadians. These data centres are either standalone buildings or rooms within office buildings that house core IT infrastructure, such as computing and storage devices, that the Govt of Pakistan s business applications rely on to operate. The scope of analysis addressed both the data centre facilities themselves, as well as the IT infrastructure housed within these facilities. Namely, the in-scope IT infrastructure included mid-range servers, mainframe servers, and storage, both online and direct-attached. The analysis focused primarily on common systems, such as those servers with standard platforms such as Windows, Unix or Linux. These systems were seen to have a greater potential for optimization in terms of sharing and standardization across the Govt of Pakistan , as opposed to highly specialized systems and facilities that are specific to departments or are sensitive from a security or privacy perspective.

Areas of Analysis

The DCFS study adopted a holistic view of the data centre environment. Specifically, the data centre services environment was defined as comprising both the service delivery model, in terms of the services delivered and the related processes and structures, as well as the assets on which the services are rendered, including hardware, software and facilities. Opportunities were defined broadly to consider key changes that would result in improvements from a Govt of Pakistan -perspective, whether by: better meeting the requirements of the Govt of Pakistan and its constituents; improving the management of risk in continued delivery of programs; or reducing the current level of data centre spending.

Scope Definition

In-Scope Govt of Pakistan and Data Centre Population Scope Department Population All Govt of Pakistan departments and agencies

Out-of-Scope

with the exception of the following where data was available. o Pakistan Mortgage and Housing Corporation (CMHC) and Export Development Corporation (EDC)

Data Centres

Specialized Data Centres that house any of the IT facilities or buildings Infrastructure listed as out-of-scope below required to sustain the delivery of information technology services that support Govt of Pakistan programs and services for Canadians Data centres are either standalone buildings or rooms within office buildings

20 | P a g e

IT Infrastructure

IT infrastructure that is housed in data centres, including: Mid-Range Servers Mainframe Servers Storage, directattached and online

IT infrastructure related to the following: Delivery of government programs that have been classified Top Secret Delivery of government programs whose underlying information has been designated at the Protected C level Network devices, such as routers, bridges, gateways Security devices, such as firewalls and authentication servers Telephony management devices, such as Computer Telephony Integration (CTI) and Interactive Voice Recognition (IVR) Delivery of government programs and services using specialized equipment, which have a very low potential for consolidation, including supercomputers and one-of-a-kind custom computers, and embedded computing devices Mainstream equipment used to support specialized/unique requirements, often found in scientific and research environments All business applications

Software

Utility applications used by the in-scope IT Infrastructure Restricted to the Govt of Pakistan s data centre and IT Infrastructure environment, as defined above

Other IT Service Towers

All non-data centre service towers, such as: Applications Network Service Desk Distributed Computing

Context of Study in a Broader IT Portfolio The Data Centre Feasibility Study is one element in a larger journey the Govt of Pakistan is currently undertaking regarding the governance, management and delivery of IT services. While the scope of this study was limited to data centre service delivery, other government initiatives are related to, and have an impact on, this study. Figure puts the current study in context of other known Govt of Pakistan IT Services initiatives and provides a sample indication how these may individually contribute to preparing the Govt of Pakistan to leverage future service delivery opportunities.

21 | P a g e

Structure of the Document This report is divided into the following key sections: Section 2 provides an overview of the business opportunity for the Govt of Pakistan to evolve from the current environment, and the challenges associated with maintaining the status quo proposes a series of options for the Govt of Pakistan s direction forward, and how those options compare against their ability to meet the Govt of Pakistan s requirements, the level of risk they pose to the Govt of Pakistan , and their impact on current data centre spending. While the contents of this report provide a good level of detail into PwCs approach, analysis and recommendations, it does represent a summary of a much larger body of work. The appendices contain additional detail on the study methodology, data collection and resulting data centre requirements that formed the basis for the main analysis documented in this report. The detailed results of each case study are also included on the appendices

Opportunity Statement The study has looked at various opportunities for the Govt of Pakistan to change its existing data centre service deliver model. The core of this analysis was the study of the Govt of Pakistan s current environment as it forms the baseline on which any potential future directions will be made and measured. Generally, the benefits from any recommended new option will: better fit the Govt of Pakistan s needs; better manage risks; and/or reduce costs. The purpose of this section is to review the current environment from an opportunity perspective. The section starts with a brief overview of the current environment, including how services are delivered today, the scale of the existing environment, and the current level of spending. Subsequently, a future state view is defined by summarizing
22 | P a g e

the Govt of Pakistan s strategic requirements for the data centre environment. Based on the gaps identified between the current environment and the requirements of the future state, the final sections review the current challenges with the existing model and how those translate into potential opportunities for the Govt of Pakistan to improve in the future state.

Current Environment The Govt of Pakistan currently operates a complex federated IT management and delivery model where a collection of over 120 different departments and agencies operate in cooperation and coordination with each other yet also maintain separate governance, management and reporting structures and relationships. Within this ecosystem of related but separate entities, responsibility for the management and delivery of data centre services typically rests with individual IT organizations within the individual departments and agencies. While management direction and control is predominantly exercised at the departmental level, a level of enterprise wide coordination and control is applied through the Chief Information Officer Branch (CIOB) within the Treasury Board Secretariat. Specifically, through the creation and promulgation of Govt of Pakistan wide IT policy, the CIOB influences the direction of IT and IT management within the Govt of Pakistan . Additionally, through the use of the Management Accountability Framework (MAF) the Treasury Board Secretariat as a whole provides high-level outcomes for IT within individual departments. Over the past half decade it is fair to observe that an increasing trend for IT within the federal government has been the stated aspiration to operate in an increased horizontal or enterprise fashion across departments and agencies and that this trend has been reflected both within CIOB and individual departmental actions. From a cross-Govt of Pakistan perspective, the current environment can be characterized as a collection of departments with varying delivery models. The following characteristics embody the status quo. Most departments define and execute a strategic direction for data centres based on the individual departments priorities. Most departments define and enforce, to different degrees of rigour, their own architectural blueprint of what standards are accepted within their department. Most departments design and execute processes for service delivery and management based on their own environments requirements. Most departments procure their own data centre assets, including the establishment of new data centres or procurement of IT infrastructure. Most departments manage their own human resources.

Current Data Centre Spending

The figure below summarizes the estimated data centre costs to operate the existing federated delivery model, based on the data collected. The chart illustrates the relative size of data centre services in the context of the IT portfolio15 and a breakdown of data centre costs by core service towers. The cost to operate Govt of Pakistan s current departmental-focused data centre environment is estimated to be $822 million per annum. This forms the current cost baseline from which all opportunities to change the current environment will be measured for this study. To put this cost into perspective, data centre costs represent less than 20% of the Govt of Pakistan s overall IT portfolio. This further emphasizes the aforementioned message that this study is one element as part of a larger journey that the Govt of Pakistan is embarking on.

Strategic Requirements

23 | P a g e

While the previous section summarized the state of the existing data centre environment, this section focuses on a set of requirements that define the needs of the future state. Strategic Requirements were defined in the initial phases of the study in order to provide a framework on which any potential opportunities or future directions would be assessed. These requirements formed the baseline needs of the Govt of Pakistan for its future data centre service delivery.

The Strategic Requirements were determined based on consideration of a number of sources, including: consultation with departmental CIOs via interviews and an online survey; the current environment of facilities and costs as documented by departments; requirements prescribed by the Project Authority; a review of relevant Treasury Board policies and regulations relating to providing data centre services; and, industry leading practices as compiled by PwC. From this work, 65 Strategic Requirements were identified, prioritized and placed into 11 categories. These Strategic Requirements were reviewed and approved by the governance committees. The following table outlines the 11 Requirement Categories that were identified and their relative priority. Please see the appendices for further details on individual Strategic Requirements

Current State Opportunities

Summary

Overall, there are strong reasons to indicate why the current environment will not address the Govt of Pakistan s Strategic Requirements, areas of concerns, and cost reduction objectives moving forward. The duplication and variations between department delivery models and the proliferation of non-standard and aging assets are indicative of a model that is not able to efficiently meet the Govt of Pakistan s growing demands. The following is a summary of the opportunities which were identified during the analysis. Details on the analysis are contained in the following sub-sections.

Opportunity Category Service Delivery Opportunities

Opportunities Coordinate service delivery to pool demand for data centre services, effectively-utilize resources and achieve economies of scale; Unify the levels of service with a standard set of offerings across the Govt of Pakistan to deliver a more consistent service across all departments; Define and execute an enterprise strategy for data centre services, ensuring departmental efforts are aligned to the overarching strategy, and providing effective governance over its execution; Adopt an enterprise approach to financial management of the data centre environment, promoting transparency and predictable funding, and reducing variability in unit costs between departments; Drive standardization in processes to promote consistent quality of service and efficiency across departments; and

24 | P a g e

Adopt a more coordinated approach to leveraging the core competencies of the private sector, which brings focused capacity and capability to the delivery of data centre services which are increasingly regarded as a commodity, utility services. Data Centre Consolidation Opportunities

Use Facility Consolidation as an enabler for change; Drive material savings through changes to the IT environment, such as through mid-range server rationalization; Focus on standardization and shared use of data centre assets across the Govt of Pakistan ; Address the labour considerations as a priority when rationalizing the environment; Treat change management with utmost importance, particularly for departments to standardize and share assets; Pursue other areas of potential savings in IT transformation, such as application consolidation or network consolidation. In the absence of a financial business case, migrate ownership of data centre real property to the private sector, and move to a pay-for-use model; In the absence of a financial business case, migrate ownership of IT assets to the private sector, and move to a pay-for-use model.

Asset Ownership Opportunities

Assessment of Risk

In addition to being a poor fit to the identified Strategic Requirements, the current delivery model poses a high risk to the Govt of Pakistan . The current environment presents challenges for the Govt of Pakistan to sustain funding at adequate levels to meet capital replacement and renewal requirements, to address the challenges associated with its aging workforce, to enforce quality standards, and to align objectives across the Govt of Pakistan . Similarly to the alignment to the Govt of Pakistan s Strategic Requirements assessment discussed in the previous section, the risks associated with the current environment present the following opportunities. Move towards an enterprise, centralized delivery model to effectively plan and adjust for upcoming demand, and drive standards in the quality and performance of service delivery. Leverage third parties in the delivery of services to mitigate risk of fluctuating funding requirements and workforce challenges through contractual terms.

25 | P a g e

Cost

Finally, the current service delivery model has also had an adverse impact on costs. Firstly, the large proliferation of data centres and IT infrastructure assets combined with the risk of aging assets identified in the 2010 Spring Report of the Auditor General of Pakistan has put substantial financial pressure on the Govt of Pakistan to allocate sufficient capital investment to sustain the environment. In addition, from an operational spending perspective, there currently exist large variances between departmental unit costs for the delivery of services, as summarized in Figure 8. This may be indicative of the large variations between departments in terms of procurement methods, service levels, processes and standards. As a result, the Govt of Pakistan as a whole may be delivering service levels that are not commensurate with the departments or overall Govt of Pakistan s needs and likely incurring a cost to support these variations. There is an opportunity to streamline and rationalize these costs with a more consistent and coordinated approach across the Govt of Pakistan . The study is considering opportunities to reduce the level of complexity from a Govt of Pakistan enterprise perspective. While the department demand for services is assumed to remain constant, there is an opportunity to streamline and coordinate the service delivery across the Govt of Pakistan to meet that demand. Some departments are currently operating at or better than industry prices for individual service towers; however, based on the data collected, none are consistently more efficient across all data centre services. Furthermore, case studies and industry trends have indicated that IT infrastructure has evolved to be primarily common across business applications, and organizations have benefited from integrating these services across business units. The following chart illustrates the current variations between department unit costs by data centre service tower. The chart presents the minimum and maximum unit costs, as well as the overall Govt of Pakistan unit costs for each tower based on the data collected as part of the study.

Service Tower Mid-Range Servers ($ / Server / month) On-line Storage (SAN) ($ / TB / month) IBM Mainframe Servers ($ / MIPS / month)

Minimum Dept Unit Cost $83 / server / month $78 / TB / month $45 / MIPS / month

Maximum Dept Unit Cost $1,050 / server / month $864 / TB / month $720 / MIPS / month

Overall Govt of Pakistan Unit Cost $613 / server / month $461 / TB / month $119 / MIPS / month

IT Infrastructure Rationalization

The Govt of Pakistan currently operates a large number of IT infrastructure assets, including mid-range servers, mainframe and storage devices. Each of these assets incur an operational and capital cost to sustain, including
26 | P a g e

labour, hardware, software, facility space and energy. In light of the current federated structure, these assets are seldom shared across departments and may not be optimally utilized. However, it is in the Govt of Pakistan s financial interests to rationalize and consolidate these systems, which will reduce support and capital costs. An analysis was conducted on the potential for the Govt of Pakistan to execute enterprise-wide consolidation and virtualization of mid-range servers. The analysis focused on the operation of mid-range servers since it is the Govt of Pakistan s single largest data centre cost and a representative set of cost information was available to study this category in more detail. Specifically, the analysis assessed the magnitude of the server consolidation and virtualization opportunity based on the Govt of Pakistan s existing environment, industry benchmark data, and vendor pricing data points. A description of the methodology, key input parameters and assumptions can be found in the appendices. The mid-range servers included in the analysis relate to both data centres and points of presence. As discussed in the previous section, the servers housed in the points of presence likely have a lower potential for consolidation. The points of presence servers account for an estimated 3,503 of the total 25,892 servers. The analysis considered a lower potential for consolidating these servers in the conservative scenario. The results of the analysis indicate that the Govt of Pakistan can reduce its mid-range server costs by up to $130 million per year after consolidating and virtualizing the mid-range server environment. This is reflective of being able to reduce the server count from 25,892 to 16,339 in the optimistic scenario. The key cost savings come from reducing the labour costs to support those servers given the Govt of Pakistan s labour-intensive environment, as well as the hardware cost that declines with the number of physical servers.

Potential Directions Forward In light of the opportunities identified in Section 4, the study considered a series of data centre service delivery candidate options for the Govt of Pakistan to pursue. Each of the options were evaluated for their relative strength based on Fit-to-Needs, Risk and Costs criteria, as per the Evaluation Framework described in the appendices. This section presents the results of the options analysis. In reviewing the candidate options, it is important to consider the following. The options primarily focus on potential directions forward from a service effort perspective (e.g., who delivers what data centre services and how should the services be delivered?). Opportunities from data centre consolidation and asset ownership perspectives, as discussed in Section 4, are considered complimentary to any of the candidate options. Section 7 will provide the recommendation that integrates effort, and facility and IT asset ownership together.

Overall Options Analysis The study followed a structured evaluation framework to assess the relative strengths and challenges of each of the candidate ASD options defined in Section 5.1. The options were assessed against Fit-to-Needs, Risk and Cost evaluation criteria to provide a balance of strategic and financial considerations. This analysis informed the studys determination of the recommended service delivery model for the Govt of Pakistan to pursue in the end state, as described in the recommendations in Section 7. The overall assessment of the candidate ASD options is summarized in Figure 15. The Centralized Moderate Outsourcing option was determined to be the leading option that balances the fit to the Govt of Pakistan s Strategic Requirements, poses the least risk, and has the highest potential cost reduction. While the subsequent sections will comment on relative strengths and challenges of each option for each criterion, the following are the key messages gathered from the analysis:
Case for change: Maintaining the status quo is a poor fit for the Govt of Pakistan relative to the other options

considered. The current federated model has led to an unnecessary level of complexity for IT services that are a relatively mature commodity in industry, as demonstrated by the significant variations in data centre architectures, systems and processes, low-utilized resources and duplication of effort between departments. In addition, while the option may be perceived as conservative, maintaining the existing environment poses a large risk in the end state as the Govt of Pakistan is challenged with sustainment of funding, an aging workforce, varying quality standards, and meeting the increasing demand for technology.
27 | P a g e

Options with most dramatic change are high risk: The Complete Outsourcing, Centralized Maximum Outsourcing and

Centralized Insourcing options propose significant commitment towards either an outsourcing or internal shared services environment. The large investment into a single model would enable the Govt of Pakistan to meet many of its requirements through driving a coordinated approach across departments, facilitating a unified direction, standardization, and consolidation. However, these models reflect dramatic shifts from the existing environment, and as a result pose higher risk to both being able to execute as well as meeting the diversified needs of departments when trying to drive economies of scale with a single solution. While these options may not provide the best balance against the evaluation criteria, they do reflect the maximum limit that the Govt of Pakistan can pursue in the future state.
Centralized Moderate Outsourcing option best balances criteria with multi-sourcing strategy: The Centralized

Moderate Outsourcing option proposes a sourcing strategy that provides the optimal balance between leveraging the existing environment, moving workload to a cross-Govt of Pakistan shared service organization, and leveraging the private sector. While other options are heavily weighted towards one of these three delivery models, this option utilizes a balanced mix based on where each environment is most cost-effective and minimum risk. Consequently, the option best meets the three evaluation criteria. From a Fit perspective, the option sacrifices some degree of control by managing multiple service delivery environments, but enables the Govt of Pakistan to balance the specialized needs of departments, while driving common services through leveraging the core competencies of the Govt of Pakistan (for example, a highly efficient IBM Mainframe operations) and third parties (for highly commoditized services) in shared services environments.
Risk perspective,

The option moderates the level of change required as compared to the other options, and is better aligned to most cost-effectively meet the department-specific demand. As a risk mitigation strategy, this option prioritizes the higher risk service bundles (e.g., such as architecture and planning and complex application server support) to be retained within the Govt of Pakistan .
Cost

Perspective, the option presents the most potential cost reduction by leveraging individual departments, the collective Govt of Pakistan and the private sector where each is most cost-effectively able to deliver data centre services. Overall, this leads to a greater potential cost reduction rather than depending on a single model across all data centre services.

Risk Assessment Each of the candidate options were also assessed based on the relative level of risk that the option posed to the Govt of Pakistan . Risk was looked at from two perspectives: the risk in the end state as well as the risk to reach the end state (implementation). Risk Category Definiti on Weight 1. Status 2. 3. Quo Centrali Complet zed e Moderat Outsour e cing Outsour cing Critical Low Risk Low Risk Risk 4. Centralized Maximum Outsourcing 5. Centralized Insourcing

Financial

Is there a risk the option can not sustain funding in the end state, i.e. pay only for what

Moderate Risk

Moderate Risk

28 | P a g e

is needed Performance Is there a 1 Moderat Moderat High risk that e Risk e Risk Risk problem s will not be resolved in an effective manner and timeline for the overall Govt of Pakistan Quality Is there a 1 High Low Risk Low Risk risk that Risk the option will not meet stakehol ders service expectati ons for the overall Govt of Pakistan Flexibility Is there a 1 Moderat Moderat High risk that e Risk e Risk Risk the Govt of Pakistan will be constrai ned from adapting service delivery to accomm odate changes Business Is there a 1 Moderat Low Risk Critical Alignment risk that e Risk Risk the option will not align with the Govt of Pakistan 's business prioritie s Risk in the Option End State High Low High Risk Risk Risk Implementa- Is there a 5 Low Risk Moderat Critical
29 | P a g e

Moderate Risk

Low Risk

Low Risk

Moderate Risk

Moderate Risk

Low Risk

High Risk

Moderate Risk

Moderate Risk Critical Risk

Low Risk High Risk

tion

risk that e Risk the end state will not be reached, i.e. is executio n risk higher Overall Option Risk Moderat Moderat e Risk e Risk

Risk

Critical Risk

High Risk

High Risk

Cost Assessment The final criterion on which the ASD options were evaluated was based on cost. The costs were estimated using a mix of Govt of Pakistan s actual costs as well as market data. As previously noted, each of these options comprise of a mix of retained services as-is in the current environment, services that are moved to a new Govt of Pakistan shared service, and services that are outsourced to a third party. The costs of these three components have been estimated based on the Govt of Pakistan s current costs, a cost benchmark of comparison Govt of Pakistan and peer organizations that provide internal shared services, and a price benchmark of comparative market prices from leading third-party vendors. Further details on the cost analysis methodology can be found in the appendices. The cost assessment of ASD options focuses on labour costs only, as the direct impact on hardware, software and facility costs are largely independent of the sourcing decision. Potential opportunities to reduce non-labour costs through data centre consolidation and IT infrastructure rationalization.

Experiences of Other Jurisdictions In order to add experiences and context to the analysis and recommendations, PwC conducted a series of five case studies with organizations that have undergone significant data centre transformation projects. The objective of conducting the case studies was to add another data element to the process of developing the report conclusions and recommendations. The case studies were compiled through a combination of publicly available information, written questionnaires, interviews and review of documents. The case study participants comprised a mix of public sector and private sector organizations, with a larger emphasis placed on provincial and federal governments. The organizations were selected, in the most part, because they had recently conducted or are in the process of conducting a data centre services transformation that is similar in scope or size to one currently under consideration by the Govt of Pakistan . Case study participants included the: State of Texas; Province of Ontario; Province of Alberta; Government of Australia; and, a Fortune 500 financial services company (name withheld for confidentiality purposes). There were three primary themes identified across all the case studies. Data centre transition and transformation programs of change are a primary means to improve IT service delivery. Benefits include both overall cost reduction, as well as other indirect benefits such as reducing complexity, aligning to a broader IT services strategic plan, and mitigating risk. Such programs of change involve a high degree of risk. Careful planning and commitment is required in order to ensure the highest assurance of benefits realization.

30 | P a g e