Recouping the Wasted 60 – 70% of IT Operations Budget

IT operations

There are 3 malfunctioning areas, but there are ample opportunities for driving improvement 

 

 

Enterprises have built up their organizations and processes over the years based upon the needs of the time. The executives were not attempting to drive best practices but were just trying to get the job done. Eventually these processes became codified and modified over time as variants to the process were required. While some organizations have gone back and instilled best practices in areas of the company, most organizations have not implemented best practices and therefore significant amounts of wasted resources and spending consume huge portions of the IT budget.

The Procurement Process

There are four areas within the procurement process that account for most of the waste: product selection; financial analysis; financing method; and lack of an overall acquisition strategy.

Product selection.

In too many cases, particularly in the area of servers, the product short list fails to include certain platforms because of an entrenched belief that the selected vendor's product must be of a particular chip, network or storage architecture. This bias then flows through the rest of the process, leading to the wrong purchase.

TCA vs. TCO.

A total cost of acquisition (TCA) analysis is good for understanding the investment costs but it does not reveal the true total costs of ownership (TCO). Business and IT executives should be making a buying decision based upon the holistic TCO analysis. For example, drawing least cost conclusions on the production costs alone of a mainframe vs. x86 servers could lead to acquiring the wrong server. The analysis needs to address the entire ecosystem – server, storage, network, cabling, people, power, floor space, etc. for the production and development/test boxes over a five year cycle (including refreshes if appropriate). When the total ecosystem is analyzed, more often than not a different conclusion is drawn.

Financing.

It is not unusual for companies to go with a financing method that is not the best for their business model. In fact, RFG finds that many companies just buy equipment outright because that is the only thing they know how to do and it is the simplest approach.

Acquisition strategy.

A company without an acquisition strategy rarely plans for and gets the best deals.

For example, if one knows the company intends to buy 100 servers over the next two or three years, negotiate a deal for the maximum and not acquire them one by one. With the right strategy, it is possible to cut the costs by 40 percent or more.

The Asset Management Process

There are three asset management process components that account for most of the waste: aging, underutilized assets; lack of optimization; and no cohesive architecture.

Aging, underutilized assets.

Aging, underutilized assets are expensive to run and from a relative performance perspective they are far from being a decent price/performer. For example, four-year old servers can consume a majority of the power while only delivering less than five percent of the performance capacity. Moreover, older x86 processors often become unstable as utilization levels reach and exceed 40 percent. This underutilization drives up costs.

Lack of optimization.

Most servers are not optimized due to a multitude of reasons, most of which made sense years ago but may not be relevant today. Standardization is lacking and the level of virtualization falls short. The average server is running less than 8 VM instances, which is inefficient, sub-optimized and costly.

No cohesive architecture.

Multiple operating systems running islands of computing still abound today in organizations. Development teams sub-optimized their decisions based on the application they were delivering, not on the optimization of the existing production platforms. So while cloud (and cloud standardization) is the buzzword for today, legacy development teams tended to build solutions that often required new hardware and software to be installed on the data center floor. Then Jerry-rigged interconnects were made to bring it together. Even with SOA today too many of the old artifacts remain that make the architecture look like a patchwork quilt. Costly to run and maintain.

Next The Lifecycle Process

Pages

About the author

Cal Braunstein

Mr. Braunstein serves as Chairman/CEO and Executive Director of Research at the Robert Frances Group (RFG). In addition to his corporate role, he helps his clients wrestle with a range of business, management, regulatory, and technology issues. 
He has deep and broad experience in business strategy management, business process management, enterprise systems architecture, financing, mission-critical systems, project and portfolio management, procurement, risk management, sustainability, and vendor management. Cal also chaired a Business Operational Risk Council whose membership consisted of a number of top global financial institutions.

Website