Why You Should Be Thinking About Demand Management

Untitled Document

Editor's note: Learn the best practices in moving data to the clouds right here.

Traditionally, CIOs have focused on a supply-side approach to IT cost control. And in large part, they have succeeded -- witness the steady and consistent declines in IT unit costs over time.

Despite these improvements, total IT spending has increased. The reason: unchecked growth of demand for IT resources.

Today, it's no longer enough to simply deliver IT resources to business customers at the lowest possible price and the highest possible service levels. Rather, CIOs must work with business leaders to define strategies to manage demand for and use of IT resources, and to raise awareness of the cost implications of different choices.

Top-performing organizations are finding ways to do just that -- through virtualization, storage management strategies, chargeback, and developing new incentives for service providers.

Why manage demand?

The chart below illustrates the divergence between IT efficiency (declining unit costs) and business use (growing volume), and underscores the imperative to improve demand management of IT resources.

While even the most elite performers can continue to squeeze incremental savings from efficiency improvements, most operations have effectively used benchmarking and other techniques to effectively manage the supply side of IT costs. As such, businesses are recognizing that the demand side of IT service delivery represents the most effective way to rein in spending, identify and achieve significant new savings opportunities, and enhance business performance.

The CIO's role in demand management should not be to direct or dictate usage levels. Rather, an effective leadership role for IT involves defining choices and quantifying the cost implications of business decisions regarding the use of IT. Put differently, the business leadership should define the priorities and forge the necessary compromises; IT's job is to implement projects and explain why some are taking precedence over others.

Virtualization as a demand management tool

What's better than a server (or a disk drive, or a network switch) that's cheap to purchase and run? A server you don't have to buy at all. In a nutshell, that's the appeal of virtualization -- the ability to meet business needs with existing resources rather than new investments.

Despite this potential, one of the challenges to implementing virtualization is to understand and quantify how a virtualization initiative delivers benefits. Part of the problem involves defining effective measures, as traditional ways to calculate the cost of physical servers are largely irrelevant to a virtualized environment. As the virtual world matures, companies are developing meaningful new metrics and improving their grasp of how the technology can yield savings.

Getting a grip on storage

Storage environments represent the area of IT perhaps most in need of demand management. Storage hardware has shown an especially dramatic gap between unit cost decline and utilization growth.

This divergence reflects a particular dynamic: when first created, data is typically placed in the highest-performance, highest-cost storage tier within the enterprise. As the data ages, the business should define various criteria and establish a process for migrating that data to lower-cost storage platforms, and eventually either archiving or eliminating it.

Instead, what often happens is that organizations manage all of their data at the platinum level, over the long term. Many CIOs have convinced themselves that the declining cost of storage hardware makes it feasible to simply throw more resources at the problem. But the exponential growth of new data that must be managed makes it clear that this approach will no longer work.

Often, applications use monolithic database structures that can frustrate the storage team's efforts to migrate data to less expensive platforms. In many cases, the best solution is for the development team to add intelligent data lifecycle management features to the applications themselves, thus giving key end-users the ability to archive, recall, and purge old data at appropriate times.

Applications may also be keeping more copies of data than necessary. To implement this type of solution, those responsible for infrastructure must identify the opportunity and enlist the business to help set priorities for development. The end result can be a very significant cost impact that goes straight to the bottom line.

Effective chargeback: the keystone of demand managment

Chargeback programs that directly link IT usage to business requirements and decisions are the most effective tool in managing demand, primarily because they introduce market-oriented transparency into the process.

In the storage example described earlier, nothing happens without the participation of the business. If the business unit pays for their storage resources based on a simple allocation linked to departmental headcount, what's the motivation to help drive change? They will pay the same, regardless of whether they use more or less.

Because using less will help the company as a whole? That argument won't get you too far in the real world. But consider: if you can say, "Since 40 percent of your storage utilization will move from Tier 1 to Tier 2, and 20 percent will be eliminated completely, the monthly storage allocation that hits your budget will decrease by 35 percent." Now the business sees clearly and directly the costs and benefits that are linked to their decisions.

Service provider relationships

Another imperative for effective demand management is to create incentives for service providers to reduce utilization. In outsourced environments, charges are typically tied to business usage. As such, service providers have little reason to help clients reduce their appetites for IT resources, since to do so would reduce their revenue.

To change this dynamic, businesses must go beyond traditional gain-sharing arrangements and facilitate an ongoing dialogue with service providers, whereby the client invests in new initiatives in exchange for the service provider's help in reducing utilization of existing services. Indeed, reducing the conflict between the client's desire to manage costs and the vendor's desire to maintain their revenue stream is the most difficult challenge in outsourcer management today.

A related approach is to provide service providers additional flexibility to implement change within client environments -- rather than putting the onus solely on the service provider to find savings. In many instances, outsourcers inherit established "ways of doing things" within the client organization that pre-date the outsourcing engagement. Giving the service provider leeway to standardize these client-specific practices can allow them to leverage economies of scale and enable lower pricing and efficiency.

New approaches

Evidence suggests that, for the foreseeable future, "doing more with less" will continue to be the mantra for CIOs of large organizations. To address this challenge, top-performing IT leaders are going beyond a focus on honing the efficiency of IT service delivery and quality, and engaging with the business to develop a variety of new ways to manage demand for and use of IT resources. These approaches involve understanding utilization drivers, implementing demand management and utilization tools, and developing incentives for the business to curb the unproductive use of IT resources.

About the Author

Scott Feuless is a Principal Consultant with Compass Management Consulting, a global firm specializing in the improvement of IT and business operations for large organizations.

More by Scott Feuless