Shared Services:
Strategic pointers to avoid the implementation fault lines

by Andrew Graham

Shared service is a transformational change in how governments operate. One might think that taking administrative units and services out of line departments and giving them back through a single service agency would be welcomed. After all, are not government departments about policy, program and service delivery? These backroom functions are secondary.

But the traditional bureaucratic organization is a vertical set of structures, within a single command and control framework, all focused on a small number of linked sets of policy and program areas. Because shared services challenge these entrenched values and modus operandi, the transition can be difficult.

The Conference Board of Canada recently held a symposium in Ottawa, and will soon be publishing on the topic. The research looked at the implementation process within public sector organizations that had already experienced some of the bumps and grinds in the conversion. It added hospitals that had built common service units and some organizations that had managed in a common service environment for many years, with many partners, so that the research could more fully examine issues of product definition, costing and governance.

The shared service model has some features that challenge the status quo and need to be understood and managed:

·     The tradition that departmental managers can only be accountable for what they actually operate. Of course, this concept – of having direct control – has evolved in many directions with the increased use of special agencies, contracting for services, third party delivery through both the voluntary and private sector. Nonetheless, it is a deeply held view.

·     The fear that ‘losing control’ of financial, personnel and related support service delivery will reduce the capacity of the organization to meet its needs. This is a genuine and real problem. At the heart of the shared service concept is a customer relationship that, for many in government, has to be learned. It is in place when the customer knows what she wants, can cost it and can measure it. The fear articulated here also reflects the reality that many governments have learned over the implementation process, that is, most departments did not really understand their needs but were content with what they had.

·     The fear that a shared service agency will become hostage to central agencies and not serve the program needs of departments. This fear is well worth holding on to. Many governments struggled with the separation of the administrative policy-making function from the actual delivery of administrative services. Shared services is only concerned with the latter.

·     The fear that shared services will cost more. In general, the research showed that costs were a difficult issue due to past anomalies among departments, the fact that many services were essentially uncosted or poorly costed and that, in a rapidly changing world of technology, costs for some services, especially those in upgrade mode, will escalate.

The Conference Board study reviewed the experiences of several governments and public sector research around the world. The field is rich with experience – possibly the richest and most aggressive efforts have been made in Canada at the provincial level. Ontario, Alberta and British Columbia have been involved in a number of forms of shared services for close to a decade. Their experience is relevant to the current federal government initiatives. Ignoring others' experience will simply add to the costs of implementation.

The Conference Board, after many interviews and much research, concluded that the effort to build a shared service environment was well worth it. However, governments need to be quite sure about why they are doing it. Marginal cost savings can be achieved in a number of ways. The traditional one is simply to take money ‘off the top’ of budget allocations. This is not what shared services is about. While some up-front savings are possible, shared services can only deliver once effective consolidation of services is in place and a client-centered business transformation process begins. Driven by economies of scale and a singular focus on administrative matters, a shared service organization can bring genuine savings and change. But it takes time.

The time perspective is important. In the public sector, there is a tendency, at least at the political level, to view the announcement of an initiative as its completion. In between is the rather messy implementation process, involving a lot of trial-by-error and experimentation. The research concluded that the process of actually implementing a shared services environment will take at least six years. Staying the course can be challenging with elections, changes of government, and changes in the project team.

There is no magic formula for successful implementation. In many respects, the most instructive parts of the Conference Board research were the stories of failed starts, restarts, and lingering issues. In virtually all cases, even those with many years of experiences and some successes to show for it, see their efforts as works in progress.

Some key elements of successful implementation emerged:
·     Verify that there is strong support for change. This cannot be a side announcement in a budget. Rather, the process of getting the decision has to be firm.
·     Ensure there is effective governance at the outset. This is a multi-party exercise. It is not something that simply happens to helpless departments. They must be actively engaged through effective governance. So too must the policy setters. Using external advisors helps as well.
·     Allocate sufficient resources to manage change over the long haul. This is not a costless transition. It is also about a major shift in how administration works within government. Change has to be resourced and managed.
·     Choose the right kind of leadership for the shared services organization. Relationship management is a key part of the leader's role. So too is organizational savvy and credibility among the key holders of influence, usually departmental heads.
·     Address resistance head-on. Who said guerilla warfare in bureaucracies does not work – it has to be dealt with because of all the opportunities for resistance.
·     Engage key stakeholders in the initiative. Beyond involving them in governance, stakeholders should be involved in evolving business solutions. This is not, however, just about having departmental representatives. That is important, but it can perpetuate the traditional vertical perspective. Specialized communities, such as finance and human resources, need to be fully involved. So too do vital partners including administrative policy experts and potential users in the broader public sector.
·     Ease the transition of staff into their new roles. While it is important to signal change, it must also take place in a caring way.
·     Build trust through service-level agreements. Such agreements are a two-way street. Often, for the first time, they force the client to articulate needs and resource commitment. For the provider, they are a source of accountability.
·     Build on process improvement. Very early in the implementation, the shared service organization has to focus on a few key process improvements to show progress and signal intent.
·     Continuously review performance against targets. The targets have to be realistic. Evaluating performance has to be visible and in the hands of both providers and clients. A common performance framework is essential.

At the heart of the research by the Conference Board is the message that this change process is harder, takes longer and is more difficult than most of those who embarked on it had imagined at the outset. They also report that it is worth it in terms of business transformation, effective cost control and the ability to innovate. Some things are just worth doing.


Andrew Graham is an adjunct professor, school of policy studies at Queen's University in Kingston. He is a senior research associate with the Conference Board of Canada and conducted the primary research for the study. He is the past president of APEX and a former ADM in the Correctional Service of Canada and Agriculture and Agri-Food Canada (grahama@post.queensu.ca or www.conferenceboard.ca).