Shared Services and the Cloud: Seize the Opportunity

The federal government has been gradually adopting shared service business models for administrative services for nearly 30 years. Today, the buzz is all about the cloud and its potential to transform shared services as we know them.

Theres much hype and a tendency to conflate shared services and cloud computingthings that have many similarities but are not exactly the same. As tips of the spear in an all-out war on government inefficiency, shared services and cloud computing could help drive hundreds of billions of dollars in long-term savings while enabling enormous transparency and performance improvements throughout the government.

Similarities and Differences

The term shared services was coined in the 1990s to describe a then-emerging business model created to transform the delivery of common administrative services, typically scattered throughout large enterprises, by consolidating them into centralized shared service organizations (SSOs). Consolidation became necessary and desirable with the advent of enterprise resource planning (ERP) software. These powerful toolsets required that the business processes they serve (primarily financial management, human resources, and acquisition and supply chain services) streamline and standardize to avoid costly customization to unique requirements in different business units.

In their purest form, shared services differ from other models of consolidated service delivery mainly in the degree of customer empowerment they enable and in their mission to operate as self-sufficient business activities through cost-recovery pricing. Customers of SSOs generally play active roles in their governance, and may take or leave their services voluntarily. Customers of other consolidated service models are often locked-in to mandatory, permanent relationships with little or no influence on service cost or quality.

First-wave SSOs usually owned and operated the technology platforms that supported their services. As they evolved, SSOslike other maturing enterprisesoutsourced more of their IT and business processes to larger-scale external providers. SSOs have thus evolved from their original roles as direct-service providers into more sophisticated roles as brokers, integrators, and deliverers of increasingly virtualized services.

Like shared services, cloud computing is a business model for delivering highly standardized services to multiple customers. The term refers primarily to technology services (infrastructure, platform, and software) while shared services connotes the entire range of back office administrative functions, including IT services, and the processes, organizations, and systems that support them. Cloud-enabled business process services are also emerging in the market; however, industrial-strength business applications and processes necessary for large scale government operations are not yet cloud-ready.

Two features distinguish cloud computing from other IT-service delivery models. The first and truly defining feature is the clouds utility-like acquisition and pricing model that enables services to be purchased like electricityon the meter as they are consumed. The other is that cloud services are always accessed and delivered electronically through the Internet (the cloud); traditional shared services are sometimes but not always Internet-accessed.

Both shared and cloud services come in a variety of deployment models. Internal SSOs serve only the parent or host enterprises in which they were created. Thirdparty SSOs are independent service providers that serve customers external to the SSO. Hybrid SSOs are internal SSOs that also serve external customers.

Similarly, private clouds serve a single enterprise; community clouds serve multiple customer organizations with similar requirements; public clouds are thirdparty providers to unrelated customers; and hybrid clouds are combinations of the others with standard technology features that enable portability.

Is the cloud transforming shared services as we know them? Not really. All shared services have been steadily evolving toward greater virtualization and Internet accessibility over the last decade. The main difference is utility pricinga distinctive and unique feature of the cloud. Tables 1 and 2 compare and contrast cloud and shared services relative to their defining characteristics.

Customer Value Creation

Given their similarities, its not surprising that shared services and cloud computing produce customer value in similar ways. Both models free customers from owning and operating their own service infrastructures or applications, thereby enabling improved focus on enterprise core missions. Both provide customers improved cost predictability and control and safer modernization paths in which costs and risks are born by external service providers. Both enable more rapid provisioning of customer requirements, thereby accelerating speed to market. As a byproduct of standardizing and consolidating processes, software, and data bases, both enable vastly improved visibility into business activities and performance management capabilitiesbenefits of growing importance throughout the public sector.

Multi-tenancy is a central feature of both shared services and cloud computing that leverages economies of scale to drive down service costs for all customers. There is another value implication relative to the efficiency and effectiveness of the governance (decision making) process in a multitenant cloud environment.

Decision rules are embedded in workflow tools within the cloud that enable routine customer change requests to be initiated and executed almost instantaneously without the need for human interactions in review and approval cycles usually necessary in noncloud environments. Electronic governance thus enables nearly real-time reprovisioning of services to changing customer requirements with improved transparency and consistency in decision making.

Shared services and cloud computing both produce significant cost savings to their customers, but in somewhat different ways. Most of the savings in traditional shared services are realized in reduced direct labor costs as business processes are standardized, streamlined, and consolidated. As SSOs mature and virtualize, additional labor savings are available through labor arbitrage. Indirect savings also accrue through reduced space, energy, and supervisory costs as service delivery footprints shrink. Technology cost savings are realized as multiple legacy systems, software licenses, and system maintenance costs are terminated and replaced by a single instance of ERP software owned, operated, and maintained by the SSO.

Cloud computing reduces costs primarily by driving out excess capacity in fixed cost investment (both fixed labor and technology costs) associated with underused IT infrastructure. Technology assets such as data centers and server farms throughout the public and private sectors often operate at considerably less than 50 percent capacity. Acquiring service from a cloud provider through utility pricing can transform a high fixed cost service requirement into a more easily managed variable cost requirement while driving down the cost of underutilized IT capacity to nearly zero.

Two decades of experience with traditional shared services has demonstrated typical savings in the range of 25 to 30 percent in functions supported by ERPs. Cloud computing has less history, but early returns are showing savings potential from 25 to 75 percent in IT services.

Potential Cost Savings

The federal government does not use business-like cost accounting practices that would enable precise calculations of potential savings from government-wide shared and cloud services transformation, but reasonable estimates can be made based on a few relevant facts. The federal government will spend $86 billion for IT services in 2010a figure that is expected to grow about 5 percent per year to $112 billion by 2015. Non-mission critical IT infrastructure most adaptable to shared services and the cloud represents about 30 percent of the total spend, or $25 to $30 billion. Its hardly a stretch to imagine $15 billion per year in savings by optimizing cloud infrastructure services to peak performance levels.

Now we get to a trickier question: How much could be saved if shared and cloud services were optimized throughout the governments entire administrative footprint? Consider these proxies from the commercial and nonprofit sectors. In most companies, administrative overhead runs are 13 to 15 percent of revenues. In the nonprofit sector, overhead is typically 20 to 30 percent of total budgets. An unusually efficient nonprofit, the American Red Cross, operates with a 6 percent administrative cost.

Many government executives believe the governments cost of administration is between 1 and 10 percent of total outlays, depending on the size of agency budgets and agency business characteristics. The U.S. Social Security Administration, one of the few agencies that routinely report administrative costs, reported an administrative cost ratio of 0.9 percent in 2009. At the other end of the spectrum, the U.S. Department of Defense (DoD) reports a broadly defined overhead cost ratio of 42 percent, of which about 16 points represent administrative functions.

Lets assume an administrative cost ratio in the non-defense sector of between 1 and 6 percent; in other words, at best the nondefense sector is as efficient as SSA and at worst its no less efficient than the American Red Cross. If so, the nondefense sectors total cost of administration would be somewhere between $26.9 billion and $161.4 billion (per Table 3). If shared and cloud services were optimized to peak performance levels enabling 25 to 50 percent lower costs across this entire base, savings between $6.7 billion and $80.7 billion could be realized.

Similarly, between $25.5 billion and $51.0 billion could be saved in the defense sector. Potential savings from both sectors combined could amount to $32.2 billion to $131.7 billion per year. The midpoint of this rangeabout $82 billion (plus or minus)might be a reasonable point estimate of savings potential across the entire government.

A recent benchmarking study of federal shared service costs suggests that 25 to 50 percent is a reasonable range of expected savings from shared services optimization in the federal environment. The study found the median cost per employee serviced by federal HR shared service providers to be 22 percent higher than the cost of similar shared services in the private sector. Therefore, 22 percent is probably a fair estimate of unrealized savings potential in sub-optimized government shared services. Significantly greater savings are likely available in areas subject to less shared service transformation to date than payroll and HR services.

A Call to Action

The governments most significant achievement in three decades of shared services gradualism has been elimination of scores of agency-specific payroll systems and consolidation into four centralized providers that serve the entire government today.

To this day, most agencies continue to self-serve for most administrative services. Redundant shadow staffs remain scattered throughout most agencies. Inefficient legacy systems continue to operate despite faster, better, and cheaper shared service or cloud computing alternatives. Most government shared services currently operating are under-used and under-performing relative to the state-of-the-art in other sectors.

The government remains stuck in an obsolete, industrial age organizational model with vast redundancies and inefficiencies. It has flat-out failed to transform with the times into a lean, high performance enterprise suitable for 21st century challenges. Why isnt the government moving faster? If we can put a man on the moon in a decade, why did it take nearly 30 years to consolidate payroll?

The answer is leadership: No president has ever declared transformation of the bureaucracy a critical national objective and invested sufficient political capital to produce more than marginal results. Congress has been a more than willing accomplice in transformation avoidance. Appropriators have shown more interest in protecting jurisdictional fishbowls than supporting transformation spanning multiple agencies and committee fiefdoms. As a result, tens of billions of dollars are wasted every year because the political establishment has failed to step up to these challenges.

The George W. Bush administration tried to push the envelope by identifying 26 ripe transformation targets through its e-government and line-of-business initiatives. After nearly 10 years of hard work, none (other than payroll) have achieved universal adoption and fully modernized peak performance.

The Obama administration has launched promising initiatives to consolidate data centers and accelerate adoption of cloud computing. Agencies have been directed to implement projects and identify savings and performance improvements in future budgets. The jury is out on whether execution and results in the agencies will meet the high expectations of the administrations leadership team.

The Defense Business Board has reported that the DoD does not routinely go to war on overhead costs and lacks effective tools to attack the tail in the departments tooth-to-tail ratio. The board recommended adoption of shared services and other best business practices as front line weapons in an aggressive battle plan.

In remarks delivered at the Eisenhower Library on the 65th anniversary of V-E Day, Secretary of Defense Robert Gates announced an all-out war on bureaucracy and a mission to shift more than $100 billion in tail costs to tooth mission activities. To paraphrase Secretary Gates: No more studies are needed; the mission and challenges are clear; the time has come for political leaders to provide the necessary leadership.

Enforcing acceptance of standardized systems throughout the government would be one of the toughest, but most critical challenges determined leaders must face. Like the tax code, government administration is rife with complexitythe byproduct of over-designed, agency-unique systems. Agencies must be forced to accept plain vanilla and give up fancy flavors with marginal business value. Moving agencies onto common platforms is fundamental to the streamlining and consolidation necessary to unlock potential savings. It would also open up the government like never before to transparency and performance management improvements.

Bipartisan legislation to codify and accelerate management reforms can be a force multiplier for their success. Achieving bipartisan support would be challenging in a business-as-usual political environment, but in todays context of a struggling economy, uncontrolled deficits and citizen outrage over government failures perhaps the time has come to make fundamental transformation a priority.

It would take a JFK-like call to action, bolder and more actively engaged presidential leadership in partnership with Congress, and a decade of hard work to capitalize on the opportunity. Hundreds of billions of dollars in potential long-term savings are on the table; tough decisions by responsible leaders following Secretary Gates example are all that is required to realize them.