Four Common Pitfalls to Avoid in Shared Services Programs

Shared services requires employee buy-in, realistic expectations.

by , / October 22, 2009
Four Common Pitfalls to Avoid in Shared Services Programs Illustration by Tom McKeith

The public sector is seeking ways to maintain its services and programs in today's difficult economic environment. To cope with unanticipated revenue shortfalls, mounting costs and increased service needs from taxpayers, governments are looking to squeeze more benefits from their shrinking budgets by adopting or expanding shared services business models.

Shared services involves streamlining and consolidating key support functions from several agencies or departments into a single, stand-alone entity, with a strong focus on customer service. The idea is to create standardized services -- using common processes, procedures and technology -- that can be delivered faster, and more efficiently and consistently.

In its most recent annual survey of state technology officials, the National Association of State Chief Information Officers found that shared services rank second only to consolidation of services and resources as a priority in 2009.

The shared services approach is attractive because it can reduce overhead and full-time equivalent costs by grouping functions together and operating in a hub-and-spoke model. However, public CIOs face significant challenges to implementing shared services, and they need to evaluate the decision very carefully to avoid four common pitfalls.


1. Misunderstanding Requirements for Federal Funds

When state and local governments use federal funds to pay for shared services, they must allocate and recover costs according to strict guidelines contained in U.S. Office of Management and Budget Circular A-87, Cost Principles for State, Local and Indian Tribal Governments.

One of A-87's purposes is to ensure that the costs of shared resources are allocated fairly and that no one misuses federal funds, intentionally or unintentionally. If state and local governments don't comply with A-87's requirements, they risk having to return millions of dollars to the federal government.

A-87 distinguishes between cost allocation systems and billed-service rates. Because most shared service initiatives use billed-service rates and not cost allocation to pay for themselves, it's important to understand the difference between the two.

Cost Allocation: A-87 requires state and local governments to allocate indirect costs according to a clear and measurable basis. For example, if several counties share an emergency dispatch service, costs can be allocated based on population and the number of emergency calls made in each county. The process is straightforward when state and local governments share one service that has one source of federal funding. The accounting becomes more complex, however, when governments receive funds from multiple sources.

Billed Services: More commonly, state and local agencies bill one another for services provided through shared arrangements. A-87 explains how to account for billed services, but it doesn't instruct state and local governments on how to develop billing rates, which can be set by any reasonable method. A-87 does require government entities to measure the profit or loss on each billed service and provide a refund or credit to the customer, or adjust future billing rates.

Agencies that accumulate more than 60 days of "working capital" from their shared services programs could be considered to be making a profit under A-87. In such cases, if federal funds were used to finance the shared initiatives, the federal government could reclaim a portion of the surplus. Agencies that don't routinely reconcile billed services to actual costs risk significant audit findings and potential refunds.


2. Regarding Technology as a Solution Instead of a Tool

In some cases, state and local governments position shared services as a technology initiative that will immediately provide better and cheaper results. Technology, however, is just a tool, not a solution. Problems are not solved by simply purchasing more powerful computers or more robust software applications.

It's crucial to examine the underlying business processes and modify or replace ineffective procedures before automating them. Automating a flawed process can actually make things worse, because now you have an expensive, new process for people to learn that still doesn't work properly. Governments must fix the process first, and use technology as a tool to streamline the workflow.

A classic example is Web services. Frequently government agencies believe that posting a form on their Web site will be cheaper to administer and easier for people to use. However, posting an application form on a Web site as a fillable PDF generates new challenges. When the applications come in, how are transactions going to be processed, reviewed and approved? What kind of retention policies are in place for Web information versus hardcopy information? What about privacy laws? Can online forms be signed electronically? Do state laws let agencies collect fees online? If not, people still have to mail checks or make credit card payments, which raises associated security concerns.

These are just some of the issues that can arise when a single agency turns to technology without thinking through the underlying business processes. The issues become even more complicated when processes are being consolidated in a shared services project.


3. Failing to Secure and Maintain Buy-In

Despite all the advantages they offer, shared services programs require a lot of buy-in to succeed. People can feel threatened by change, and shared services programs are all about change. The change that comes with shared services can be widespread -- affecting jobs, responsibilities, budgets and even the number of people a manager supervises.

Advocates for shared services need the support of supervisors, peers and subordinates for their programs to succeed. Otherwise, people who are threatened by change could sabotage the efforts to provide improved and more cost-efficient government services.

A good example of the need for buy-in typically occurs in information security. When state and local governments look to develop shared services models that pool IT resources, law enforcement agencies sometimes resist because they feel they need to control access to their data files. For instance, they might resist co-locating their servers in the same data center as the health department because their information is so sensitive.

In reality, there's no technical reason why law enforcement servers can't reside next to other departments' servers. In this case, demonstrating that law enforcement servers will have the appropriate level of security is vital to achieve buy-in.

In shared services models, employees must know that leadership at all levels supports the initiative -- from governors to mayors. The support of these top executives must permeate all levels of government. If it appears people are undermining the program, governors or mayors must let people know that the shared services initiative is more important than any one individual's job security. Within an organization, line managers must reinforce this message.

Service-level agreements can help maintain buy-in by detailing responsibilities and costs between service providers and their internal clients, and by providing a framework for measuring deliverables according to predetermined schedules. If you can't prove that implementing shared services is a better way to operate, then the program won't be well received.


4. Expecting Too Much from ROI Projections

Implementing technology for shared services programs generally won't reap immediate savings. Instead, technology implementations typically cost more in the first few years than they save.

Technology requires investment -- in hardware, software, training, even real estate, like if a new location for the data center has to be found. Moreover, when transitioning to a shared services model, government agencies usually run parallel systems for three to six months or more to ensure that the new system can deliver what's needed. In addition, assumptions about recovering costs by selling old equipment don't always pan out. Many states prohibit government agencies from selling used computers for privacy reasons.

All of these factors indicate the need for accurate return-on-investment (ROI) projections. However, ROI projections for shared services models are a moving target. Until a shared services program has been established for two or three years, expenditures and savings can vary for many reasons. Government agencies chart a new course when they launch shared services. Ultimately it gets them where they want to go, but there can be detours, delays and unanticipated problems along the way, all of which can influence the projected ROI.

The best use of ROI projections in shared services is to identify the main categories of expenses and savings -- the "buckets" for hardware, software, training, consulting and transition costs. Analyzing these buckets can generate a number of questions: Why do we think we are going to save money? Why do we think we're going to save time? Why did we spend 20 percent more than anticipated? How can we stay closer to budget next time?


Achieving Success With Shared Services

Shared services business models can yield significant savings while maintaining or even improving service delivery for state and local governments. However, the path to success contains many pitfalls that can snare the unwary. Agencies and departments can avoid these traps by understanding the rules for using federal funds, regarding technology as a tool, achieving buy-in and understanding the limitations of ROI projections.

Michael Claytor Contributing Writer
Michael Claytor, certified public accountant, is a partner with Crowe Horwath in Tampa, Fla. He can be reached at 813/209.2438 or
Geoff DePriest Contributing Writer
Geoff DePriest, project management professional, is a senior manager with Crowe Horwath in Indianapolis. He can be reached at 317/208.2552 or