Lessons Learned from the Transformation of Pay Administration Initiative

Government of Canada model for accountability

Under the Westminster model of government, deputy heads are professional, non-partisan public servants positioned at the head of departments and agencies. They are selected by the Prime Minister on the advice of the Clerk of the Privy Council, and appointed by the Governor in Council through an Order in Council. Deputy heads serve as expert advisors to their respective Ministers, and are in charge of day-to-day management of their department on behalf of the Minister. Deputy heads are accountable to multiple parties.^{Footnote 18} They are accountable to their Minister in the exercise of their duties, but also to the Prime Minister through the Clerk of the Privy Council on matters that are of importance to the collective management of the government. Deputy heads are also accountable to Treasury Board for the management and performance of their respective departments^{Footnote 19} and to other central agencies with regards to the powers entrusted in them (which include management of human resources). However, in this model of accountability, a degree of ambiguity remains with regards to who is responsible for the implementation of government-wide initiatives.

More specifically pertaining to people management, deputy heads have primary responsibility for the effective management of the people in their organizations. They are responsible for planning and implementing people management practices that deliver on their operational objectives and for assessing their organization’s people management performance. One of the people management practice principles is “An effective people management information infrastructure supports business success and accountabilities.”^{Footnote 20}

As a central agency, TBS is the administrative arm of the Treasury Board, and the President of the Treasury Board is the Minister responsible for TBS. TBS supports the Treasury Board by making recommendations and providing advice on program spending, regulations and management policies and directives, while respecting the primary responsibility of deputy heads in managing their organizations, and their roles as accounting officers before Parliament. The Treasury Board oversees the government’s financial, human resources and administrative responsibilities, and establishes policies that govern each of these areas. The Treasury Board, as the management board for the government, has 3 principal roles. It acts as:

• the government’s Management Office by promoting improved management performance
• the government’s Budget Office by examining and approving the proposed spending plans of government departments and by reviewing the development of approved programs
• the human resources office and employer or People Management Office by managing compensation and labour relations for the core public administration^{Footnote 21}

Organizationally, there are 3 broad areas within TBS, namely the Office of the Chief Human Resources Officer (OCHRO), the Office of the Comptroller General (OCG), and the challenge function (including Program Sectors and policy centres such as the Chief Information Officer Branch (CIOB)). OCHRO’s responsibilities include the management of compensation, including compensation planning and reporting, collective bargaining, establishing the collective agreements, establishing the common business processes for people management, and the management of HR-related functional communities (with the  compensation advisor community as a particularly relevant community for the transformation of pay administration). The OCG’s responsibilities include monitoring, providing guidance and recommending corrective actions regarding financial management performance of departments and agencies and providing advice on control frameworks for financial management (with compensation being a large component of the operating expenditures of government^{Footnote 22}). The challenge function or due diligence of TBS involves cost validation, assessment of risk, alignment to current priorities and sound advice on submissions in the overall role of general manager of the federal public service.^{Footnote 23} The challenge function draws heavily upon the various Treasury Board policy areas of the Government of Canada including not only Information and Technology, but also areas such as Risk, Assets and Acquired Services, Financial Management, Compensation, and People Management.^{Footnote 24} The Information Technology Project Review and Oversight Division (ITPROD), a group within CIOB at TBS, has a mandate for the oversight of a select portfolio of high-risk or complex IT-enabled projects within the Government of Canada.

In terms of service provision, under the Treasury Board Common Services Policy, PSPC (formerly PWGSC) is identified as a common service organization with compensation being identified as a mandatory service for which departments and agencies must use PSPC as the common service provider.

In terms of accountabilities regarding compensation, the Department of Public Works and Government Services Act requires the department to provide such administrative and other services required for the disbursement of pay, employee benefit plans and superannuation or pension plans to persons employed in or by any department, and to persons employed in or by other portions of the federal public administration, as the Governor in Council may direct. Moreover, as directed by the Act, by Order in Council, the Pay Disbursement Administrative Services Order (2011) repeals the 1981 Order and states (in part) that the Minister of the department shall provide services required for the disbursement of pay including the verification of payment requisition certification, the development, operation and maintenance of a pay processing system and take all necessary steps with respect to pay administration to initiate, change or terminate pay based on personal employee information and pertinent personnel documentation.

Cost saving measures

Around the time the TPA was launched, the government was actively pursuing a range of cost-reduction objectives. The 2010 and 2011 Budgets reflected an intention to reduce government spending, namely by having Treasury Board carefully manage department growth and freezing operating budgets to 2010 to 2011 levels. The 2011 and 2012 Budgets, building on 4 rounds of reviews initiated in 2007 to improve management of government spending, introduced a 1-year Strategic and Operating Review spearheaded by Treasury Board (in 2011) and a comprehensive Deficit Reduction Action Plan (in 2012). The goal of these measures was to achieve at least $4 billion in ongoing savings by fiscal year 2014 to 2015 (or 5% of the review base).^{Footnote 25} The Strategic and Operating Review covered 67 departments and organizations representing about $80 billion of direct program spending.^{Footnote 26} The Deficit Reduction Action Plan also entailed generating ongoing savings of 5% to 10%^{Footnote 27} of direct program and operations spending through increased efficiencies in departments and agencies.

Other contextual factors

At the time TPA was launched, each department managed its own technological resources and related budget, which resulted in a decentralized and uneven IT landscape across the public service. In order to remedy the obsolescence of IT systems in the public administration, the Government announced in 2011 the creation of Shared Services Canada (SSC), a new organization which was charged with modernizing, standardizing and consolidating IT resources across 43 departments. The initiative aimed to reduce costs by updating the technological structure and centralizing IT staff and budgets within SSC. SSC’s broad mandate includes reducing the number of data centres from almost 500 to 7,^{Footnote 28} consolidating and securing 50 overlapping networks, and ensuring all emails are managed through a single common infrastructure, among other commitments. SSC also equips departments and agencies with updated hardware (for example, voice over the internet for phones, video-conference, wifi).

The Human Resources Services Modernization (HRSM) initiative was launched by TBS in 2011. The objective was to modernize human resources (HR) services in departments to reduce the number of HR systems to support 1 process, 1 system and 1 data set for the Government of Canada, leading to efficiencies for transactional HR activities. The HRSM initiative included several components, but of particular relevance to TPA was the implementation of Common HR Business Processes (CHRBP). The CHRBP, led by OCHRO, was meant to “bring consistency to the delivery and management of HR services in all government organizations.”^{Footnote 29} There is no reference to alignment with pay systems or processes in the foundational CHRBP documentation. The implementation of CHRBP entailed that organizations would analyze and report on their departmental HR practices based on the new business process, and would then identify, prioritize and implement actions to ensure their alignment with the CHRBP. This synchronization process was deemed by OCHRO to be completed by March 31, 2014. However, an evaluation of the HRSM initiative in 2016 found that, while all participating departments and agencies indicated they had implemented CHRBP, a few of those consulted for the evaluation did not feel their organization had fully adopted CHRBP and that their organization continued to use previous processes (especially pertaining to the roles continuing to be carried out by HR personnel rather than managers, with the establishment of “shadow” HR units).^{Footnote 30} Moreover, the evaluation raised concerns that the CHRBP was not implemented to a sufficiently detailed level in order to ensure government-wide consistency for IT-systems and data and found that there was a continuing need to implement processes at a more detailed level.^{Footnote 31}

My GCHR, introduced in 2015, is the Government of Canada-branded version of a PeopleSoft 9.1 Human Resources Management System, which, with onboarding taking place over many years^{Footnote 32} is meant to replace the over 70 different HR systems that were in use across the Government of Canada.^{Footnote 33} The synchronization to this new platform is meant to increase automation, standardize and streamline HR processes across departments, increase self-service options, and facilitate information sharing.^{Footnote 34} As of June 2017, My GCHR was only introduced in 44-organizations and there are an additional 31-different-HR systems remaining in departments and agencies.

Appendix C presents the landscape of the various HR-systems that operated within the Government of Canada at the time of the launch of Phoenix in 2016. It illustrates the various systems and interfaces required of the new pay IT-solution at launch.

Scoping: Properly define what is changing

A Target Operating Model construct as illustrated in Figure 2 indicates the scope of areas to be considered in any transformation initiative.

Figure 2 – Text version

This figure consists of a series of hexagons that is based on a model for the finance function developed by KPMG. There is one large hexagon in the background. Centred on each of the six sides of the large hexagon’s perimeter is a smaller hexagon. There is one small hexagon in the centre that contains text. Each hexagon on the large hexagon’s perimeter contains text that describes the six aspects of a model for the finance function.

The hexagon in the centre contains the text “Federal Pay Operating Model.” Centred above it, the first smaller hexagon contains the text “Services, Functions and Processes.” Moving in a clockwise direction, the next five hexagons each contain text that describes another aspect. They are “Technology,” “Organization and Governance,” “Clients,” “Performance Management” and “People and Skills.”

While each of the areas in this model was actually undergoing change as part of the TPA, the focus was almost wholly confined to the management of technology (the Pay Modernization project pertaining to development a new IT solution) and of organization (the Pay Consolidation project pertaining to the downsizing of the compensation advisors and establishment of the Pay Centre).

However, the initiative implicated much more than these 2 aspects. The TPA Initiative was complex, broad, and highly dependent on the ability of a wide range of users to prepare for the transition, and adapt and change the way they carried out their HR and pay activities. While TPA documentation frequently referred to the complexity associated with the processes and business rules for pay, most of those consulted for the study admitted that very few people (other than compensation advisors) understood the degree of complexity associated with the day-to-day requirements to ensure accurate pay. The complexity of the HR to Pay process and how it would be transformed may have been understood in segments or silos but the overall management of the end-to-end process transformation from, for example entering HR data for an initial hiring offer, to the regular entry of overtime or other compensation actions, to paying employees and benefit providers (through an overall process owner^{Footnote 36}) did not appear to be in place.

Business processes, people/skills (including compensation advisors, others working in HR and managers with HR responsibilities), all users of the new IT solution (that is, all public servants), organizational culture, services and functions (for example, the transforming of the roles of managers and HR professionals in departments and agencies), quality and timely provision of HR data and related HR systems were also very significant areas of change in this initiative. While some of these other aspects were mentioned in the early documentation (PWGSC’s Report on Plans and Priorities for the 2010 to 2011 fiscal year that briefly discussed business processes), it was only in passing and no comprehensive plan for transformation was implemented.

Cutting across all these elements was the establishment of a new service model for the administration of pay. Where before departments and agencies had compensation advisors internal to their organizations, these positions became centralized at PSPC for 45 departments and agencies.^{Footnote 37} However, the design of this service model (including culture, service standards, roles and responsibilities, governance, processes) did not address the full scope of the changes end-to-end across HR to Pay that would be required within individual departments and agencies, nor did the development of the service model fully address the diversity of practices within all departments and agencies.

To further illustrate the complexity of the environment within which the Government of Canada administers its pay and the resulting impact of this environment on the performance of pay under RPS, a benchmarking analysis undertaken by IBM compared the Government of Canada pay administration performance with industry performance in the areas of cost, efficiency and quality, and cycle time.^{Footnote 38} This study confirmed the complexity of the environment within which pay is administered in the Government of Canada and suggests that this complexity should have been addressed (at least in part) prior to implementing a new IT solution.

Thus, additional discrete projects (scoped and funded and with assigned accountability) were needed to address all aspects of change per the operating model. Some examples of these additional projects include:

• implementing new business processes (and controls) in departments and agencies to be carried out by those in HR (including compensation advisors) and by managers (such as CHRBP, but to a greater level of detail and fully implemented by departments and agencies)
• implementing changes to existing HR systems and introducing new HR systems (such as My GCHR)
• implementing an HR to Pay performance management framework, including benchmarking key service, efficiency and quality performance metrics for the original pay administration environment so that the performance of the new pay administration environment using Phoenix (by all departments and agencies) could be properly understood and tracked
• addressing the cultural aspects at play such as the resistance to changing long-established practices such as carrying out most HR transactions retroactively, which has contributed to substantial backlog of pay transaction (as outlined below under Lesson 9)

Roadmap: Articulate discrete projects and how they interrelate

When considering implementing a complex initiative, a roadmap is a high level description (which can be a graphic, a table or narrative) of the various components of the projects, along with the schedule, deliverables and interrelationships.^{Footnote 39} Leading practice suggests that a roadmap should be developed that is based on the discrete components that fully account for the proposed changes to all aspects of the operating model, as well as related projects.

The roadmap encourages project managers to see how each part of the project (or in this case, each discrete project in the TPA Initiative) ties to:

• initiative objectives
• timeline
• deliverables
• risks
• interdependences and interrelationships

The roadmap is a tool whereby each of the discrete projects would then be considered holistically. This would include not only project management, but when and how the concepts of change management and risk management would be brought in. It is the view of the study team that if the various discrete projects had been considered as a whole, the interrelationships and interdependencies would have been identified and could have been better managed and associated risks better mitigated.

Even in the absence of the identification of all the various discrete projects, a more holistic and inclusive approach to project, change and risk management would have addressed many of the issues encountered throughout the TPA lifecycle. Indeed, if a roadmap had been outlined, then a better staging of the various components could have been undertaken. The study team suggests that the staging of the components could have been:

• the business processes within PSPC and also within departments and agencies pertaining to HR to Pay (including roles and responsibilities with respect to approvals) should have been examined and changed/fully implemented first. This would have included training managers on their HR responsibilities and any of their changed roles, including those pertaining to Section 34^{Footnote 40} of the Financial Administration Act
• next, new or updated HR systems would have been implemented (ideally based on a common configuration of PeopleSoft across most departments and agencies) and the quality of HR data addressed in a systematic way in all organizations
• this would have been followed by the build of a new pay system based on the new business processes and on the common HR system implemented, along with the training on the new system of an experienced compensation workforce. Reduction of the number of compensation advisors could be managed based on efficiencies within the new pay system
• the last part of the transformation would have been the relocation of compensation advisors to a more centralized delivery model

Accountability: Assign accountability and authority to a single office

Accountability is a relationship based on obligations to demonstrate, review, and take responsibility for performance, both the results achieved in light of agreed expectations and the means used. The principles of effective accountability are outlined below.^{Footnote 41}

From the perspective of roles and responsibilities, as noted previously in Lesson 1, the overall undertaking within the Government of Canada to transform the compensation environment was shaped by reducing of the scope of the transformation to be simply the 2 projects under the TPA. Accountability for both projects under the TPA (that is, Pay Modernization and Pay Consolidation) was assigned to PSPC. As noted previously, its roles and responsibilities as the overall lead were sought by the Minister of Public Works and Government Services in 2009, and granted, following the normal challenge process of the TBS.

However, as noted in Lesson 1, the actual scope of the undertaking to transform the compensation environment, which involved much more than these 2 projects, was not well understood. One result of the too-narrow scope was an incomplete accountability framework for the transformation. The roles and responsibilities of additional parties with accountabilities for the broader areas (namely TBS along with departments and agencies) were not effectively designed, documented nor implemented as part of an overall accountability framework.

An important party in the transformation of the compensation environment was the deputy head community responsible for leading affected departments and agencies. Deputy heads are accountable for their own organizations and therefore, for the implementation of TPA within their organizations. However, the study team learned that most deputy heads were not aware of most of their responsibilities with respect to the TPA, particularly ensuring new business processes introduced by the CHRBP initiative were fully understood and implemented, and also implementing change management more broadly. Moreover, most deputy heads did not realize the extent of the changes being introduced by the initiative and the corresponding risks to their organization and to their employees’ pay. Reasons for this varied but the most commonly mentioned factor was that there were many change initiatives underway at the time and TPA was not identified to deputy heads as being more important than any other. Also, briefings provided by PSPC throughout the initiative were generally positive and, other than raising concerns on the need to ensure HR data was accurate, did not raise any red flags or other reasons to warrant more deputy head attention.

Another principle of effective accountability is that of balanced expectations and capabilities; that is that performance expectations should be clearly linked to and balanced with each party’s capacity (authorities, skills, knowledge and resources) to deliver. PSPC sought and was granted the accountability to carry out both projects on behalf of the Government of Canada. However, it did not appear that the authority to ensure that the full scope of changes to occur throughout the whole of the Government of Canada was either recognized as being needed nor was it given to or assumed by any one organization. So although PSPC had the accountability to deliver on TPA, neither PSPC nor any other organization was able to exercise the necessary authority to make this happen.

With PSPC not being able to exercise the overall authority (neither at the deputy minister nor ministerial level) over the full scope of the transformation of the compensation environment, the senior coordinating body for the transformation success appeared to fall to the Public Service Management Advisory Committee (PSMAC).^{Footnote 42} PSMAC regularly received updates on TPA and even served as a sounding board for deputy head concerns about the project; however, the committee could not be accountable for the initiative. No one individual or governance body was assigned the authority to ensure the necessary steps were being taken to ensure the overall success of the initiative.

In a government-wide initiative such as the transformation of compensation, it is clear that an effective and comprehensive accountability framework should explicitly include the authorities and roles of TBS, departments and agencies, as well as of any organizations leading specific projects within the initiative. The accountability framework should also clearly state where the overall accountability rests for the full scope of the transformation initiative. This overall accountability should be specified at both the ministerial and deputy head level.

The actual selection of the appropriate minister, deputy head and organization for overall accountability is complex and would likely depend on the particular transformation initiative. At the moment, it does not appear that any one organization has the necessary accountability, authority and capabilities to carry out such a role in future transformations. Given that these could be adjusted, one possibility to consider would be TBS, given its “general manager” role.^{Footnote 43}  It could assume the overall accountability for such government-wide transformation initiatives. Other possibilities could be a lead client department/agency or the department/agency leading the major project(s) within the initiative. In each case, accountabilities, authorities and capabilities would need to be addressed.

Governance: Establish broad and inclusive governance

Governance refers to how an organization makes and implements decisions.^{Footnote 44} Governance is complex and fluid and complicated by the fact that it involves multiple actors who are the organization’s stakeholders. They articulate their interests, influence how decisions are made, who the decision-makers are and what decisions are made. Decision-makers must consider this input and are then accountable to those same stakeholders for the organization’s output and the process of producing it.^{Footnote 45}

According to documentation received from PSPC during the lessons learned study, the Pay Modernization project included the following governance mechanisms:

• Senior Project Advisory Committee (SPAC)
• Risk Management Oversight Committee (RMOC)
• Executive Management Team
• Change Agent Network
• External Advisory Committee
• TPA Union Management Committee (TPAUMC)

The initiative’s governance (and decision-making) was led by PSPC but did not reflect the full scope of the areas undergoing change in the initiative. In particular, most governance was focused on the Pay Modernization project rather than the Pay Consolidation project. Documentation from PSPC indicates that the Pay Modernization governance bodies (SPAC and RMOC) received Pay Consolidation updates starting in fall of 2013. Thus, there was really no overall governance body for the TPA as a whole. Such an overall governance body, established at the outset of the initiative, should have had participation from PSPC, various players at TBS (such as OCG and OCHRO) and departments and agencies.

One of the key gaps in the governance structure was the lack of decision-making involvement of an HR to Pay process owner. Leading practice would suggest that there should be a defined accountability for the end-to-end process, in this case the HR to Pay compensation process and that this process owner should be part of governance. The SPAC and RMOC, being only advisory bodies and being focused almost exclusively on the Pay Modernization project did not fully consider the HR to Pay process. While all those consulted for the study agreed that PSPC was not the HR to Pay process owner, there was little consensus regarding who the process owner was, whether OCHRO or OCG. Regardless, while being members of the SPAC, OCHRO and OCG rarely attended meetings.

Leading practice would see external service providers (that is, firms) included as observers/members in governance bodies. According to documentation received, an External Advisory Committee was struck to provide strategic advice and guidance to the Minister and Pay Modernization project executives. It met 8 times and included 2 to 3 external members representing private sector companies. This committee received project updates at their meetings and provided advice to the project team.

The Change Agent Network was the product of combining the Empowering Change Network (established for the Pay Modernization project) and the Change Agent Network (established for the Pay Consolidation project). The purpose was to share information with departments/agencies, which could then be used to brief up and throughout the organization. It was composed of representatives of all departments and agencies affected by Pay Modernization. While organizations were encouraged to appoint individuals at a manager level or above, this was not the case for many organizations. In fact, some organizations appointed people at a much more junior (for example, AS-02) level. Additionally, attendance by many departments and agencies was uneven with some organizations hardly attending at all and other attending frequently. Generally, attendance by departments and agencies was good until October 2015 when attendance dropped to less than 50%. However, the change agents consulted as part of the study (at all levels) felt that they did not completely understand their role (as the primary point of contact between PSPC and their organization) and that information provided at meetings was not clear or adequate to allow organizations to fully understand the status of the projects or their role, particularly pertaining to change management.

The TPAUMC was established in early 2011 based on a request from the Public Service Alliance of Canada (PSAC) for meaningful consultation on the TPA with a view to minimizing the adverse effects on employees. The committee met at least quarterly from 2011 to the end of 2015. While no one consulted from PSPC or TBS mentioned this committee during the consultations for this study, PSAC spoke about the committee at length. While the union felt the committee was a very useful forum for discussing the workforce adjustment impact of the TPA on existing compensation advisors, the union did not feel the Government of Canada representatives were responsive to the concerns raised by the union at the committee about the rollout of TPA. According to PSAC, PSPC did not share key documents with the committee that would have likely led to red flags early on in the process. Additionally, according to PSAC, PSPC never wavered from its plan that the Pay Centre could be run by 550 employees, despite numerous concerns raised by the union at meetings of the committee.

While not a formal part of TPA governance, over the timeframe of the initiative, the project executives presented to PSMAC 7 times. According to many of those consulted for the study, PSMAC was often used as a senior governance body for the initiative. That is, PSMAC was asked to determine departmental readiness for go-live and to provide advice about key project milestones. However, members of PSMAC, having only received a limited number of briefings and not being completely attuned to the various initiative (and project) nuances, were not well-placed to provide this kind of advice. It is the opinion of the study team that the use of existing enterprise-wide advisory or information sharing bodies (such as PSMAC) should not be regarded as a substitute for governance of such significant transformation initiatives.

Oversight: Ensure oversight provides sufficient challenge

Oversight is a critical function in any transformation initiative, group of projects or project. It is carried out on behalf of the accountable leader in order to monitor risks, issues, schedule, budget, scope and outcome achievement. It is a challenge role, having a degree of independence from the ongoing management of the initiative or project itself.

As mentioned, TBS CIOB ITPROD has a mandate for the oversight of complex IT projects within the Government of Canada. One important way that ITPROD carries out its oversight role is through attendance as an ex officio member at project governance meetings. Membership at assistant deputy minister (ADM)-level SPACs is the ADM of CIOB, the Executive Director of ITPROD, or the assigned Oversight Executive, depending on the project. At the SPAC for TPA, the Government of Canada Chief Information Officer (that is, the ADM of CIOB) was ITPROD’s representative at the committee. However, the study team heard that ITPROD’s influence was arguably weak during the timeframe of the initiative due to cutbacks as a result of the Deficit Reduction Action Plan and the tendency to delegate attendance at SPAC meetings (according to meeting minutes, the ADM CIOB attended less than 25% of meetings). A review of the minutes from the SPAC meetings further revealed that delegated representatives rarely voiced concerns or questions whereas the ADM CIOB, when in attendance, would frequently make contributions to the meeting that resulted in action items. The study team heard that while senior managers at ITPROD did voice concerns about the dashboards, these concerns were not reflected in updated dashboards and PSPC project management leadership team took steps to reassure ITPROD and CIOB at TBS that the project was on track (for example, through phone calls).^{Footnote 46} Further exacerbating the challenges pertaining to the role of ITPROD, their role was limited to oversight of the Pay Modernization project only.

Thus, the initiative was operating within an environment of limited oversight other than the management hierarchy at PSPC. Since this oversight was within the organization, there was limited independence of those tasked with the challenge role. Within that hierarchy, the study team heard from those consulted from both within and outside PSPC that there was a culture (at PSPC in particular, but also to some extent within the Government of Canada more broadly) that is not open to risk and does not reward speaking truth to power. Additionally, we heard that this culture of discouraging briefings with bad news was particularly strong at PSPC during the timeframe of the initiative. According to many of those consulted for the study, project briefings provided a more positive report on project progress and minimized areas of concern raised by others within PSPC (including individuals from the Chief Information Officer Branch and Finance and Administration Branch of PSPC). Those who observed the practice could not say definitively whether project management executives were unaware of the problems, unwilling to admit there were problems or were rather hoping that they would be able to address the problems in a timely manner/before the next briefing.

This practice of not providing briefings that contained bad news was exacerbated by a tendency to accord a great deal of leeway to managers with a good track record of managing projects. In particular, the study team heard from multiple sources that if certain managers were assuring others that problems would be resolved or that the project was on track, senior managers/decision-makers were more likely to believe those managers rather than to seek independent confirmation of what they were being told about problems.

Given that TPA was managed in a cultural environment that did not reward those who spoke truth to power (that is, to identify and communicate problems) and a tendency to believe those with good reputations, a strong oversight/challenge function was needed. Senior management, decision-makers and those asked to provide input should have been willing to ask for an independent assessment of what was in the briefings. For example, independently validated assessment for government-wide readiness to proceed with the go-live launch (from both PSPC as well as departments and agencies) should have been sought by PSMAC when this body was asked to provide advice on readiness to launch.

Other stakeholders within the government system should have also played a stronger challenge role. For example, TBS in its role pertaining to ensuring the quality of Treasury Board Submissions and adequate follow-up and reporting to Treasury Board could have required more frequent and formal reports to Treasury Board regarding project progress, coupled with evidence of progress. Also, others with responsibilities pertaining to compensation (including OCHRO and OCG) could have stepped forward and asked more questions. Members of the various governance/advisory bodies who received briefings could have asked more pointed questions and, upon not receiving a satisfactory response, could have escalated their concerns to their deputy head. Internal audit units could have conducted audits of departmental readiness for the transition to Phoenix and/or could have escalated any concerns they may have had to the Office of the Comptroller General. Those within OCHRO could have better advocated for the compensation advisor community regarding the complexity of their role and concerns about the ability of the Government of Canada to operate effectively after go-live. The design of the Pay Consolidation project combined with the decision to locate the Pay Centre in Miramichi effectively meant that the pre-existing knowledge and capacity of the compensation advisor members was eliminated. It is important to note that compensation advisors were represented by their union, PSAC, which spurred the development of the TPAUMC. However, the work of that committee does not appear to have influenced the outcome of the TPA in any meaningful way. Decisions were made despite the warnings issued by the union representatives at TPAUMC.

It is the view of the study team (based on leading practices) that third parties who have extensive comparable experience globally should be brought in as a regular part of the challenge function to provide independent assessments of progress of the initiative as a whole and to share leading practices from other jurisdictions.^{Footnote 47} The scope of these independent assessments should be sufficiently broad and report directly to the senior governance body. While 3 independent reviews (linked with 3 of the Gates, per guidelines for project management oversight) were conducted for the Pay Modernization project, these reviews unfortunately were very limited in scope, and although following Treasury Board policy direction for IT project oversight, were not effective. A closer examination of this work reveals that those consulted for the reviews were limited to project management personnel within PSPC and that the reviews were very IT-focused. Thus, although appearing to be effective reviews, they did not identify significant problems or issues with the project, or the broader TPA Initiative and each review recommended proceeding to the next phase of the project.

Culture/leadership environment: Assess and then adapt or manage the culture and leadership environment

Organizational culture is defined as the shared values of and behaviors uniquely common to an organization. Organizational culture is integral in determining how tasks are completed, the way people interact with one another, the language they use when communicating, and the attitudes, goals, values, and leadership behaviors that are exhibited.

As they embark on a large transformation, leaders need to assess the organizational culture to determine the cultural elements within the organization that may help or hinder the change direction and achievement of expected benefits.

This assessment determines if the organization’s current culture, structure, processes, and performance management system will support the change. If so, then the assessment process will investigate how the current culture can be managed through the change. If the current culture will not support the change, then the assessment identifies the aspects of the current culture requiring change in order to realize the future state.

Indicators that a cultural change may be necessary to support and sustain the change include:

• the current culture does not allow stakeholders to work in ways that support the future state
• the current culture does not support the planned organizational process or behavior change
• the current values are in conflict with what will be expected of the stakeholders and leaders^{Footnote 48}

The public service culture and environment at the time of the change initiative was described as being unreceptive to inconvenient feedback. The study team heard that this unreceptive culture was further compounded by the existing top-down approach to management within TPA. We heard this tendency to ‘bury bad news’ and to only brief up the good news was 1 of the reasons some of the major concerns were not raised, were not considered, or did not reach senior levels.

As described in previous sections, there was no individual or governance body assigned the authority to make sure key steps to ensure the overall success of the initiative were taken. The study team heard this was further compounded by problems with the existing culture which is not open to risk and does not reward those who share negative news. As a result even if negative news was shared within a department (and it was not always done), the negative feedback may not have reached executives or decision-makers. Negative news did not necessarily travel across departments. Individual groups focused on carrying out their individual responsibilities, without realizing that the initiative’s success was dependent on the success of everyone involved.

The team also heard that there was not enough focus on ensuring buy-in and ownership at the department level. PSPC took on the role of pushing for the implementation of the IT solution while departments and agencies received the information passively and felt it was being imposed on them.

When assessing the extent to which individuals in an organization would be amenable to sharing feedback that might not be positive with their organization, it is useful to consider the advantages and disadvantages of so doing. From the feedback received during the study, the disincentives to express concerns outweighed the advantages. Employees and contractors who delivered messages of being on track, on time, and on budget were favourably received, while those who expressed their fears, worries or concerns indicated that their concerns were dismissed.

In a culture where errors and failure are not tolerated, learning and innovation are stunted and with this the agility to respond to a complex and changing transformation agenda.

Therefore, assessing the organization’s culture as well as its capacity and readiness for change (focusing on required behavior changes, skill sets, competencies and potential barriers to change) is essential for successful change and transformation.

Change management planning and execution: Treat change management as a priority

According to leading practice, successful change management requires buy-in from those affected by the change, and support for staff and managers to help them adapt to changing roles. When undertaking complex changes in an organization currently facing many changes (as described in Lesson 1 and the Context section), the use of Change Agent Networks can be especially beneficial.

Change leadership is also critical to success. Change leadership requires leaders to take both accountability and responsibility for the change, that is, accountability for outcomes management and responsibility for driving the change operationally as they attend to the people, process and technology components of the transformation.

Change leaders need to be trained to use effective strategies to facilitate organizational change and overcome resistance to change. They must be versed in change management practices but also model openness to the mindset, behavior and culture required to support change.^{Footnote 49}

Initially the TPA was on track to follow leading practices in change management. For example, comprehensive and thorough change management plans were developed early on in the initiative and these identified numerous potential risks and challenges. Their existence suggests that there was a good understanding that the TPA required a strong focus on change management to mitigate the many identified risks to success. However, as mentioned earlier, the study team heard that these plans were not fully implemented since they were declared out-of-scope when costing for the project identified higher overall costs.

As mentioned above, most of those consulted for the study admitted that very few outside the compensation advisor community understood the day-to-day requirements to ensure accurate pay. We heard that, given that pay transactions were carried out by compensation advisors who are lower-ranking employees, decision-makers may have under-estimated their role and under-valued their expertise. Underestimating the required skills, knowledge and expertise to administer pay led to a commensurate underestimation of the value and importance of the change management processes necessary to support the change.

A further challenge was the fact that the public service as a whole was undergoing a number of changes in parallel with the implementation of the project. It is a leading practice when engaging in ambitious change processes to secure the enthusiasm of those involved to help propel the change forward. In the context of TPA, immediately preceding the launch, public servants’ enthusiasm for change may have been dampened by less than positive press coverage about the value of public servants, cuts to staff and funding and increased workloads and numerous other change initiatives impacting the way they worked. The TPA was introduced in an environment of instability and fatigue. Challenging environments require more, not less, of an active change management process to ensure success. Despite this, change management for the TPA was decentralized and delegated to departments/agencies in fiscal year 2012 to 2013 as part of cost savings for the overall initiative. Representatives of departments and agencies consulted for the study reported they had insufficient resources to fully understand and implement change management.

The TPA required a significant change in the way individual employees and managers carried out their day-to-day activities. The change required that pay and compensation personnel, HR personnel, supervisors and managers perform HR and pay operations online and in real time. This meant, for example, that managers had to approve hours worked before or right at the end of the pay period rather than after the end of the pay period, which removed the previous flexibility to adjust data after the fact. Delays in entering and approving information in the system could have a direct impact on the accuracy of employees’ pay.

Many of these changes were not well understood by those affected by the transformation because of scoping issues outlined in Lesson 1. Change initiatives with such a significant human impact should have magnified the need for a comprehensive and sustained change management function and the development of change leadership competencies, particularly for managers.

The use of Change Agent Networks is considered a leading practice in change management, in particular when the change affects a number of groups and organizations. A team of change agents plays a crucial role liaising between the working level and senior management. Their role is to facilitate the engagement of system users and prepare them for the challenges ahead. Again, a leading practice with change agents requires them to have superior knowledge of the systems and processes that are being transformed, and thus serve to provide an early warning system when there are anticipated or unanticipated issues with, for example, resources, expertise, integration, functionality, processes, funding, capacity, acceptance to change.

As noted previously, a Change Agent Network was created. However the feedback provided to the study team showed that while departmental/agency colleagues appreciated the support of the change agents, these agents could not provide the full level of value as they routinely lacked current and complete information and context to share with others. We also heard that the change agents were originally at the director general level but it was often a more junior person attending the meetings that were organized. One of the rationales given by interviewees was that meetings were focused on detailed procedural information that was not adapted to a senior executive audience.

Although change agents received briefings, attended workshops and had access to information and resources, the majority did not have a broad enough understanding of the initiative, the risks and the impacts of the changes. If change agents had been provided current and complete information and context, then they could have more effectively briefed their own senior managers and also would have understood the importance of relaying their own organization’s readiness back to the PSPC project management team in a timely manner. Thus, it is critical that change agents be provided with the full scope of the change and the implications if they are not able to carry out their expected duties.

A well-managed change management function also has an added advantage of bringing together disparate parts of the change process into a cohesive perspective. In the case of the TPA, an effective change management process would have allowed for monitoring of progress from a horizontal multi-function, multi-department/agency perspective. Had this been the case, the seriousness of the issues that were not addressed (for example, as presented throughout this document, in particular departmental readiness, lack of accurate and timely data, functionality of the IT solution) because they were HR to Pay issues, could have been identified and raised early on.

Another important role of a change management function is to help all key stakeholders keep their focus on the ultimate goal. Those consulted for the study described a culture where the primary focus was more on the immediate goal of launching on time and within budget. The immediate goal was achieved successfully at the expense of a sustainable solution. A centralized change management group could have potentially identified the effect of this pressure to deliver on time and within budget, and worked to find the time, resources, and engage in discussions in the appropriate forums to ensure a renewed focus on the longer term goals.

As evidenced in the previous paragraphs, when major transformation processes are being carried out in a cost-savings environment, the investment in change management (that is, the cost to inform, educate, engage and train the key stakeholders, including users of the IT solution) becomes even more critical to the success of the initiative, and should be treated as an indispensable investment in success, rather than an optional or add-on to the change process that can be under-funded and under-utilized. Successful change management plans are integrated to all aspects of the major change, funded adequately, and change agents are given continuous and priority access to senior levels to ensure the change can be successfully carried out. Had change management plans been carried out as planned, benefits could have included: improving alignment of resources and decreased costs; minimizing disruption to day-to-day activities while preparing for the transformation; providing a forum for concerns; escalating risks which were identified; anticipating challenges; improving morale; engaging key stakeholders; increasing opportunities for information sharing across groups and departments/agencies; and leveraging best practices.

Communications: Focus on communications that are effective

Communications are an integral part of change management processes. When implementing communications strategies, the measure of success is when the conveyed messages are understood, discussed, and result in necessary actions. The more significant the change, the greater the imperative that communication be carried out successfully.

Effective communication was critical for this initiative as it involved significant change in processes, interrelated systems, and involved a large number of individuals across a myriad of departments and agencies and communities of practice.

A key principle is that the measure of success of communications is not (just) the volume or frequency or avenues of communication, but rather the extent to which the communication results in desired actions for all parties involved. Effective communication is dynamic (two-way) and focuses on exploring new ideas and adapting messages to improve outcomes.

From the start of the TPA, communications from PSPC with departments and agencies was described to the study team as a one-way, top-down process. Whether it was in-person meetings (for example at committees), or email transmissions, the information was viewed as being instructions and updates, without a mechanism or venue through which to ask questions, seek clarification, provide feedback or express concerns.

According to documentation and many of those consulted for the study, PSPC provided regular, periodic updates to various existing governance bodies as a mechanism to share information; these bodies did not form part of the initiative or projects’ governance and did not have an oversight role. These included the Executive Project Oversight Committee, the Human Resources Council, Back-Office Transformation committee and PSMAC. Ad hoc presentations, upon request, were also provided to the Chief Financial Officers and Deputy Chief Financial Officer community and the National Joint Council. Again, the study team heard that most of these briefings were one-way (that is, from PSPC project executives to the committee) and often did not include a question/answer period. If questions or concerns were raised, the study team heard that there was no mechanism or commitment to provide answers or clarification.

The study team heard that communications were further complicated by an apparent unreceptiveness to departments and agencies who were seeking discussions and negotiations on serious issues such as the processing of employees with irregular hours. We heard that client departments and agencies encountered resistance when they requested meetings with the PSPC project management team.

The study team heard that the content of the messages communicated by PSPC often lacked context; hence, the target audience did not have a full understanding of what was required of them and they did not have the background needed to help guide them when making important decisions about their own activities. For example, we heard departments received instructions to clean their HR data; however, some ended up cleaning the wrong data or not cleaning in the manner specifically required for the introduction of Phoenix. Sometimes the messages communicated were viewed as unhelpful and inconsequential, while leaving more pressing questions unaddressed, for example how Phoenix would interface with other HR systems.

The timing of communication was highlighted by many of those consulted for the study as being a key problem that contributed to ineffective messaging. Often, when the information was received it was out-of-date or inaccurate because it did not reflect more recent changes to the IT solution, processes or requirements which were evolving on a continual basis. The study team heard that communications were often delayed due to the approval process required to release these. A different governance structure (Lesson 4) would have accelerated approval.

A variety of factors, as identified through the consultations, may have influenced this pattern of communications. For example:

• the study team heard that an important driver for the project management team was meeting established launch dates; hence any feedback that might slow down or delay progress was unwelcome and resisted
• we heard content of the communication was largely focused on specific technical issues and requirements, for specific users, and the broader context was lacking. As well, we heard that information provided to decision-makers was incomplete. As a consequence the information was not available to assess the broader HR to Pay risks that were arising or the organizations’ readiness
• the study team heard that there was a lack of clarity about where users could address questions and comments, which caused some confusion and uncertainty
• finally, the study team heard from PSPC that once the communications were crafted, further communications within departments and agencies were deemed to be the department/ agency’s responsibility, which could explain why communications were not tailored to departmental realities and why the impact of communications was not measured

The IT solution

Phoenix is based on Oracle’s PeopleSoft 9.1, a COTS IT solution. In order to ensure the IT solution meets the needs of the user, it must be configured and customized. In an IT-enabled business transformation, the IT solution is but 1 component in the overall initiative. Although it is often an expensive and high profile component, its relationship with business process change and other functions is critical for success.

Functionality: Engage and communicate with stakeholders regarding anticipated IT Solution functionality

The study team heard that many departments and agencies raised concerns about the IT solution’s ability to meet the pay needs of their organization. We heard that the PSPC project management team did not provide satisfactory reassurances that their needs would be fully incorporated. When some needs were not incorporated in the end, the study team heard that the mitigation strategies proposed did not address the concerns in a timely manner. Having said that, there were a few examples where strategies were put in place by PSPC to address concerns, such as introducing the role of “timekeeper” to enter HR information needed for accurate pay on behalf of employees who were unable to access a computer in a timely manner.

Of particular relevance to explain the large degree of backlog at the Pay Centre when the IT solution went live in 2016, is that Phoenix works best when HR transactions are entered before they take place and is extremely inefficient dealing with retroactive transactions (that is, HR transactions that have taken place in the past). However, the Government of Canada’s culture and business processes are firmly entrenched in retroactive transactions. Moreover, based on consultations for the study, those within departments and agencies were not aware that pay accuracy and overall throughput for the compensation function would be negatively affected by the continued culture of retroactive transactions, despite communications from PSPC to this effect.

For example, at go-live, when a transaction related to “acting” had to be entered retroactively (which was most of the time), compensation advisors were required to make a manual adjustment or workaround. This contributed to the very large number of pay transactions requiring the personal attention of a compensation advisor (and continues to impact the backlog of pay transactions at the Pay Centre). The study team heard that there was minimal engagement of the compensation advisor community to test and assess the true cumulative impact of such a large number of workarounds.

Compounding the problem, many pay procedures were not clearly articulated at go-live. In fact, the study team heard that the complete user manual was not ready until after go-live (and it continues to be updated). Also, procedures were simply not in place for some transactions until recently (such as leave with income averaging) resulting in compensation advisors across government guessing how to do it, choosing different approaches, many of which were not correct and caused further errors in pay.

As mentioned earlier (under the Initiative Definition section), proper initiative definition and scope would have identified the mismatch between Government of Canada culture/processes regarding retroactive transactions and endeavoured to adjust culture/business processes prior to go-live to minimize the number of retroactive transactions.

Customization: Do not conduct extensive or risky customizations if they can be avoided

PeopleSoft 9.1 is a COTS IT solution that is recognized around the world. The study did not find compelling evidence to suggest that a different COTS solution would have been more appropriate for the Government of Canada. However, 2 significant challenges in implementing this COTS solution in the Government of Canada were raised that were unique to the Government of Canada and were noted as very significant challenges.

First, the decision^{Footnote 50} to split PeopleSoft into 2 IT solutions (Phoenix and My GCHR) introduced significant complexity. Experts consulted for the study confirmed that splitting PeopleSoft in this way has not been done anywhere else in the world, to their knowledge. The usual implementation would use the full PeopleSoft “out of the box” integration between compensation and HR functions, as opposed to severing all these interfaces and then re-implementing interfaces between the resultant different configurations of PeopleSoft. Complexity was increased even further by having multiple HR systems besides My GCHR continuing to operate and interface by different means.

The other significant challenge was that the system integrator (that is, the private sector entity that was responsible for configuring and customizing software products) was required to make extensive workarounds within PeopleSoft 9.1 to meet the untransformed business practices and accommodate the complex business rules. As mentioned in Lesson 9, 1 of the major challenges with the Phoenix implementation was to adopt customizations to deal with the very extensive workload of transactions that occur retroactively. This business practice, with over 50% of transactions being at least 60 days in the past, is unique to the Government of Canada and required not only customizations to PeopleSoft, but also manual workarounds. Leading practices confirm that the greater the degree of customizations and workarounds to the system, the greater the risks and erosion of the value of purchasing a world-recognized COTS IT solution. Thus, the business processes and operating model should be reviewed and amended to best leverage the availability of the functionality of the COTS application prior to adopting an IT solution.

Outcomes management

Outcomes management, value management or benefits realization are similar terms for the explicit management and tracking of the various outcomes and benefits of projects and transformation initiatives. Success must be defined and a plan to measure progress established. Performance should be assessed before, during and after implementation (including people, processes, HR data and the IT solution). Success should not be declared too early but rather be based on the achievement of outcomes.

Determining success: Implement outcomes management throughout the life of the initiative

There are 2 ways to look at the expected outcomes of the TPA. The first is to identify what outcomes that a new pay IT solution and consolidation of compensation advisors would afford the government. In the 2011 to 2012 Report on Plans and Priorities for PSPC,^{Footnote 51} several outcomes (including cost savings) were identified, including:

• Strategic outcome:
  • Government of Canada being an employer of choice
  • increased contribution to public service renewal/Management Accountability Framework
  • increased stakeholder confidence in payroll administration function
  • supports recommendations of report of the Standing Committee on Government Operations and Estimates
  • greening government operations
• Final outcome:
  • increased pay administration sustainability
  • increased Government of Canada operational efficiencies
  • increased quality of reporting for departments
  • created simplified compensation advisor organization
  • eliminated need for departments to keep compensation advisors, space, A-base, etc.
• Measured outcome:
  • increased employee satisfaction
  • decreased time to train pay administration staff
  • decreased cost to deliver pay administration
  • decreased materiel usage (for example, paper, mail)
  • decreased work effort for compensation advisors
  • increased pay administration staff retention
  • increased number of accounts per compensation advisor

The study team did not see any further evidence of these outcomes being tracked or measured or mentioned in subsequent documents. Additionally, through consultations for the study, we learned that no further actions were taken to measure or track the achievement of these outcomes. It appears that the comprehensive outcomes management plan developed in the original business case was never implemented.^{Footnote 52}

The second way to look at the success of the initiative is determining the extent to which the initiative was delivered on scope (that is, functionality of the IT solution), on budget and on time. In this definition of success, the outcomes and benefits are all implicit in the achievement of the scope. However, the study team heard that the project management team at PSPC was primarily focused on the project budget and timeframe. This meant that the scope was decreased when time and budget pressures were felt.

Even if the areas highlighted above had been explicitly measured, a number of indicators that the initiative would be successful were still missing. These include, for example, progress in transforming business processes, training the workforce on new processes, measuring efficiency and quality of service provision as changes are made, training and productivity of managers and employees in interacting with the pay IT solution, and development of and access to a complete knowledge base for the new IT system and the transformed processes, roles and responsibilities (this was called the User Productivity Kit, an Oracle Development Platform) for and by all user types.

Timing of benefits: Do not expect savings until well after implementation

As mentioned in the context section, the total amount of savings was to be $79.1 million (later decreased by $1 million to $78.1 million in 2012) per year, starting in the first year of Phoenix go-live, fiscal year 2016 to 2017. This amount was to be recouped largely from workforce adjustments of over 700 compensation advisors. Savings were anticipated to come from 4 main areas:

• savings from consolidating pay services, due to a greater degree of standardization and economies of scale ($10.8 million/year)
• savings due to seamless integration between the HR PeopleSoft system and the new pay IT solution ($35.3 million/year)
• savings due to self-service as a result of employee and manager self-service capabilities ($17.6 million)
• savings due to process changes and automated payroll calculations, as well as internal efficiencies found in ongoing operations in pay administration ($14.4 million)

Because the project planned for these savings in the first year when Phoenix went live (starting in April 2016), over 700 compensation advisors received workforce adjustment notices in October 2014. To ease their workforce transition, many compensation advisors began to seek out new positions resulting in a shortage of compensation advisors prior to go-live. The end result was that there was an existing backlog of pay administration tasks that had to be entered into the IT solution at the time of go-live that were transferred to the Pay Centre.

The expectation that 100% of annual savings could be realized in the first year was based on an assumption that not only would the new compensation advisors be working at full productivity, but that managers, employees and HR professionals would also be sufficiently familiar with the new IT solution self-service capabilities and aware of their own responsibilities, so as to be fully productive and not require assistance from a compensation advisor (or other person within HR). However, leading practice suggests that full savings could not be achieved simultaneously with the implementation of a new IT solution. The most realistic scenario would actually be an increase in costs in the first year, allowing for some compensation advisors to stay on in their organizations to support managers and employees as well as allowing for a gradual increase to full productivity at the Pay Centre. From there, cost savings would be realized gradually over time.

A comprehensive implementation of the outcomes management plan with tracking of metrics related to areas such as the effectiveness and extent of business transformation, the efficiency and quality of operations by the evolving community of compensation advisors (with departures, relocations and new recruits), and the extent of training taken and retained was essential but never done. It likely would have clearly illustrated that the financial benefits promised would not be achievable as scheduled and that quality, timeliness and service levels would be at significant risk.

Support to affected stakeholders: Support affected stakeholders if they are to play a role in the initiative

A key component of a business case is the complete and realistic picture of costs and benefits representing the full scope of the transformation initiative and the departments and agencies involved. In addition to making funding available to stakeholders who are part of the transformation, other supports in the way of communications and training are necessary for success.

The role of the departments and agencies was central in contributing to the success of the TPA Initiative. During the planning phase this role was recognized, and sound and detailed strategies for supporting and engaging the ‘people’ part of the change process were devised, consistent with leading practice. These comprehensive plans, however, were not fully implemented and the study team heard it was likely due to a lack of human and financial resources.

Specifically, in the case of the TPA, departments and agencies were asked to do more to support the success of TPA within their organization but they did not receive additional supports to undertake these additional tasks either in the form of additional human resources, training or funding. The risks were high, as their participation in the change process (such as cleaning HR data on time, training line managers on their HR-related roles and responsibilities) had a direct effect on the success of the new pay IT solution. These risks were also not fully appreciated since the scope of the project was not well understood (Lesson 1).

The lack of funding to support the new process and procedures at the departmental level, coupled with the lack of communication, poor engagement of managers in the change process, insufficient training (these issues are discussed in other sections), meant that departments and agencies were not equipped to hit the ground running when Phoenix was launched. The study team heard that setbacks may have been avoided by having a solid centralized process in effect as well as having an individual or governance body assigned the authority to ensure necessary steps were being taken. This authority could have developed strategies to ensure that the activities for the overall initiative were optimized in light of a decreased budget, and that managers across departments and agencies received necessary resources (that is, information, money and human resources) to allow them to fulfill their roles.

Testing: Fully test the IT solution before launch

The testing of software for an IT solution is a complex and lengthy process. It includes unit, integration, system and acceptance testing, characterized by increasing levels of complexity. Testing ensures that not only the IT solution is working on its own, but that it is properly interacting with other components or systems, that the new IT solution meets specified requirements and that it meets the needs of users.

Evidence from all sources consulted for the study confirms that Phoenix was not fully tested. While over 16,000 test scripts were executed that demonstrated that Phoenix could process most pay transactions, either through automation or workarounds, testing was not done in such a way as to identify the likely challenges to be encountered due to issues such as: complex procedures required to effect workarounds; imperfect or late information provided to Phoenix from HR systems; and multiple pay issues for many individuals.

A report^{Footnote 53} was issued the month before Phoenix went live for the payroll for 120,000 employees. It listed 4 types of outstanding defects, the nature of the resolution and the planned date for the fix. The 4 types of defects were related to:

1. the 10-day rule (that is, determine eligibility for allowances that require an employee to work 10 business days within a given month);
2. late reporting (retroactive) functionality defects (effective dates prior to the current Pay Period);
3. integration (between Phoenix and Government of Canada HRMS, including My GCHR and web service); and
4. defects resulting in additional manual steps in existing procedures. In all, there were 69 defects, including 19 that were major, 46 that were moderate and 4 that were minor. Temporary solutions for most were compensation advisor workarounds and fix dates for most were later in 2016 or at a time to be determined.

In the opinion of the study team, there was a large number of defects that were major and a large number of defects that had no planned fix date only a month before the first launch.

While the original implementation plan was to conduct a pilot (with Natural Resources Canada), this was cancelled because testing was not yet complete and thus the IT solution was not ready for a pilot. Also testing should have included a more extensive set of real-life end-to-end scenarios using actual users trained only on available training material with the full spectrum of what managers, employees and HR professionals would experience. Additionally, testing different types/levels of complexity of pay records with the broad spectrum of users should have been thoroughly performed but was not. The volume of testing and thus completeness and representativeness from a statistical sampling perspective was decreased in the final stages due to time and budget pressures.

Sandbox environments (that is, isolated testing environments) were either not available or not available for a sufficient period for the real go-live IT solution. This is at odds with leading practice. Reasons for this were put forward as budget pressures, time pressures, inability to stabilize the IT solution and/or difficulty in Government of Canada provisioning the actual IT environments.

While client acceptance testing was conducted (running 207 scripts with 6 departments), it was not end-to-end. Documentation reviewed for the study revealed that there was a disconnect between expectations of PSPC and departments and agencies in terms of the test process and department and agency responsibilities. As well, the study team heard that only some types of pay were tested by departments and agencies, usually not the most complex types of pay (such as seasonal, part-time, acting). Also, there was little engagement of departments and agencies in testing whereby the actual “trained” users (that is, managers and compensation advisors) involved could interact with the system to provide approvals and conduct manual workarounds. This kind of testing would have allowed PSPC to assess not only the IT solution’s capabilities, but also the capability of the Government of Canada overall to carry out the function of pay administration.

Engagement of the private sector: Leverage and engage the private sector to maximize initiative capacity and capabilities

Private sector entities were hired to configure and customize the software solution (IBM, the system integrator), provide resources to support project management (ADGA) and provide specific assistance for detailed tasks related to business or technical issues (PwC).

The division of responsibilities between PSPC and the contracted entities (including the system integrator, the project management office and advisory services) did not appear to fully leverage expertise nor take into account capacity. For example, PSPC was the project manager for the initiative as a whole and individuals were brought into certain roles through a staff augmentation contract with ADGA, as opposed to PSPC contracting a firm to carry-out work to achieve outcomes related to project management. In addition to overseeing the work of the various contracted entities, PSPC was responsible for communications, testing and training.^{Footnote 54} The study team heard from multiple stakeholders (who were also closely involved throughout the initiative) that PSPC did not have the capacity or capability to carry out the full extent of their duties. These stakeholders pointed to poor decisions, lack of detail in communications and poor quality training as indicators that PSPC did not have the necessary capacity and/or capability to carry out their responsibilities for communications, testing and training.

We also heard that the private sector entities that were engaged on the project were not encouraged to contribute to the broader initiative of TPA. This included not communicating or collaborating with others involved in the initiative and not being invited to the governance table(s), except to receive or provide briefings. Despite their expertise and experience in similar transformations, TPA was then not able to benefit from a collaborative environment with the private sector entities which would have enabled leveraging their broader perspectives pertaining to the nature of the risks and mitigation strategies, change management approaches and impacts of incompatible business processes with PeopleSoft 9.1.

A significant portion of the involvement of private sector entities was limited to the role of staff augmentation, rather than as advisors. Specifically, in project management, the contract was only for individual contracted resources with specific skills, not for the delivery of an effective project management function. Thus the private sector could not adjust and augment resources to address gaps in project management effectiveness, but rather was only accountable for staff augmentation (the provision of individuals with specific skills related to a project office).

In examining the role of the private sector, it is evident that a traditional procurement approach was adopted for the system integrator, with the government spending a number of years developing the approach and a full set of very detailed requirements (several hundred pages). The system integrator was not involved in this initial project scoping but instead were only able to subsequently take those requirements and provide a design for the new pay system including what could/could not be automated; what could/could not be integrated, etc.; they were not given the leeway to question fundamental project or initiative assumptions. One result of this was the fact that there were reportedly (from several different sources) over 2,500 change orders to the tasks included in the contract between PSPC and the system integrator. Additionally, change orders related to business requirements continued throughout 2015, impacting the timing of finalization of the final IT solution configuration, which in turn impacted the timing and thoroughness and effectiveness of documentation, testing, training materials and training delivery on the actual IT solution configuration.

Leading practices in developing and selecting private sector partners and their roles in achieving a transformed service involves much more than traditional procurement approaches with lengthy and detailed specifications developed by government. Approaches such as Commissioning,^{Footnote 55} Outcome-Based Contracting and Joint Solutions Procurement^{Footnote 56} allow the private sector and government to work together prior to settling on roles and involvement of the private sector. Under these models, the private sector may be able to assume a much larger role in designing and delivering the actual services, or the system integrator would be involved with government in designing the project and roles, fully understanding the outcomes to be achieved and the capacity and capabilities of the parties involved.

Mid-initiative adjustments: Reassess, learn and adjust

Large multiyear initiatives will likely have their overall context changed in some significant way through the life of the initiative, ranging from a change of government priorities, new or changed initiatives affecting the initiative in question and also the increased experience and understanding of the challenges and opportunities within the initiative itself.

As far as the study team could determine, the business case and outcomes management plan were not reviewed at any point during the initiative implementation. While plans were updated in 2012 following the project definition phase of the initiative, very little was actually changed. In particular, the budget was adjusted down from $192.1 million to $186.2 million and annual cost savings were adjusted down from $79.1 million to $78.1 million due to better estimates. But the harvesting of the savings was still expected in the year the new system would go live. Though the overall savings did not change, the source of the savings did change significantly as the anticipated functionality of the IT Solution was better understood. This should imply a revised outcomes management plan, for example, which was not included in 2012 updates.

Illustration 1 – Text version

This figure shows an equal-sided triangle. At each apex, the three aspects of project management are labelled: “Cost” on the top, “Time” on the bottom right, and “Scope” on the bottom left.

Project management is often seen as the managing of the inter-relationships between budget, time and scope (that is, functionality). This inter-relationship is commonly referred to as the project management triangle or triple constraint. In much of the literature on project management, the scope dimension is linked with quality. Alternatively, quality is described as the ultimate objective of the project and thus does not appear on the triangle.

The study team heard that there was resistance to request additional time (beyond the 2015 to 2016 fiscal year) and budget for the initiative. It is not clear to the study team why this was the case aside from the previously mentioned tendency to minimize briefings with bad news. During interviews, PSPC representatives indicated that they did not think they could get additional time or budget and thus did not make the formal request. However, the study team heard that there were other initiatives that successfully returned to the Treasury Board at the time to seek additional funding. The rationale for not seeking additional time is somewhat clearer when considering that compensation advisors had received their workforce adjustment notices that were effective April 1, 2016, and cancelling or further delaying this action would have contravened collectively bargained workforce adjustment provisions. While it is not impossible to seek exceptions to these provisions, it is nevertheless a major decision that must be taken at the highest levels of the bureaucracy. Given the positive reports from the project management executives, coupled with department and agency reports they were ready for go-live, seeking additional time through an exception to the workforce adjustment provisions did not seem warranted.

Thus, while the planned scope was decreased over the timeframe of the initiative due to budget and time pressures, the overall impact on departments/agencies, the Pay Centre and initiative outcomes were not explicitly reassessed. While the study team heard that the project team did make some adjustments to their assumptions about case load per compensation advisor after launch, we did not find evidence that the project team changed their messaging to departments/agencies. In particular, with the loss of many self-service and automation functionalities, there would likely be a greater reliance on compensation advisors as well as managers within departments/agencies to interact with the system in a timely manner. We could not find evidence that this was communicated to users of the system. Also, the plan for the Pay Centre to be staffed with 550 compensation advisors was never adjusted despite changes in scope and concerns voiced by the community and the bargaining agent. This would likely have an impact on the overall outcomes of the TPA, but again, we did not find or hear any evidence that impacts in departments or agencies were being factored in as part of the considerations.

Capacity: Identify and establish required capacity prior to go-live

In the case of the TPA Initiative, a large number of highly skilled compensation advisor positions were lost through cuts of their position, or their decision to seek alternate employment rather than moving to the consolidated Pay Centre in Miramichi. OCHRO was responsible for the HR functional community, including the 2,500 compensation advisors that were part of the community at the start of the initiative; however we were told the focus was on workforce adjustment more than on community management. The compensation advisor community underwent a high number of changes, in addition to having to work with new software. Many compensation advisors had to adapt to moving from being embedded in a department or agency to working within PSPC. This pressure was compounded with the loss of over 700 positions and a move to Miramichi. As a result, many compensation advisors with experience and expertise in payroll were either lost to attrition or chose not to move to Miramichi.

While staff cuts may have been motivated by urgency in achieving planned initiative savings, in hindsight there was overconfidence in the abilities of the new pay IT solution, and a serious underestimation of the necessary role of the compensation advisors in ensuring employees are paid accurately and on time.

Underestimating the required capacity to carry through the transformation to the steady state may also have resulted from the misunderstanding of the complexity of administering pay under RPS. While TPA documentation frequently referred to the complexity associated with the processes and business rules for pay, most of those consulted for the study admitted that very few people (other than compensation advisors) understood the degree of complexity associated with the day-to-day requirements to ensure accurate pay. The study team also heard that some project management executives dismissed concerns by saying that the Pay Modernization project was simply “replacing a giant calculator.” This did not accurately reflect the complexity of pay nor the critical role of compensation advisors’ knowledge in pay administration, and may have led others to assume the upcoming transformation was not as multi-faceted and complicated as it was. These misunderstandings likely led to poor decisions about required staff capacity after go-live.

Additionally, when new compensation advisors were brought on board at the Pay Centre, many did not initially have the required level of knowledge to administer pay. The study team heard that it can take several years for someone new to compensation and pay to fully grasp the work given the complexity of the federal public service pay environment driven by myriad ever-changing union agreements, and system rules and regulations.

Furthermore, no formal mechanisms had been established for the transfer of knowledge and capacity building prior to eliminating the compensation advisor positions. All involved were faced with a steep learning curve with a system which was in flux (it was evolving at a rapid rate), while having limited support in terms of tools (including user guides) as well as how to do their jobs at launch. This resulted in challenges for both the new staff and for existing staff as they were still learning the system when it was live and had real-life impacts.

Line managers had an important role to play in the new pay system, which for many of them involved taking over some of the responsibilities previously delegated (unofficially) to administrative personnel and compensations advisors. Some of these activities were time sensitive. If approvals in Phoenix were delayed, this would have an adverse impact on an employee’s pay (because manual workarounds were required resulting in extensive delays). Many line managers were caught unawares when it came time for them to take on these duties they had never exercised in a self-service environment before and which had become much more complex than in the past. The capacity of managers to complete their duties had not been adequately considered, communicated, negotiated, and addressed. Also, managers with a large number of direct reports, found themselves in situations where they lacked the time to enter the required information for their entire team, to the new deadlines.

Training is one of the foundational aspects of change processes, as it provides key stakeholders with the knowledge, information and skills needed to change the way they carry out their activities and to adapt to new responsibilities. The consistent feedback received during the study indicated that the training provided was insufficient. As mentioned above, compensation advisors did not have the information and tools to use the new Phoenix system effectively, and line managers either lacked the training (less than 50% participated in the training) or the training received was inadequate (that is, did not explain that some tasks may be new for some managers, did not explain the importance of timely approvals), and/or the time delay between the training and the requirement to put the training into practice was too long.

While the reasons for the inadequate training were not confirmed, there were numerous suggestions:

• not enough importance was assigned to training by providers (to ensure training was current and relevant and useful) or recipients (to make the time to seek and attend training)
• the pace of change was unfolding so quickly training was overlooked at times
• there were so many changes to Phoenix as it was being configured for the Government of Canada, that when training was delivered it was already outdated and not relevant
• much of the training delivered was paper-based or based on screen shots, which was too far removed from the experience of working with the actual IT solution to be of value to participants

The experience of the TPA highlights the imperative of considering and planning for sufficient numbers and skilled capacity to support change processes, in particular when the change is major and complex and relies on subject matter experts, such as compensation advisors, and a high number of users (including managers and employees).

While this lessons learned study did not explore the time post-launch, it is obvious from our consultations that no detailed contingency plan was available for implementation in the event that something went wrong with the launch of TPA. Having detailed and operational mitigation strategies in the event that some of the identified risks are realized is a critical part of being ready for the launch.

Thus, success routinely hinges on having the right numbers of people, with the right skills, sufficiently trained to take on their new responsibilities, as well as having a sufficiently detailed contingency plan that can be readily implemented if necessary.