Intel, the world's largest computer chip maker, operates a total of 44.6 million sf of facility space in more than 45 countries, 22.7 million sf of which is manufacturing space. Since 1998, Intel's Corporate Services (CS) organization has been developing an increasingly sophisticated approach to analyzing facility management practices and their associated expenditures. Using a Web-based internal benchmarking system and other collaborative tools, Intel generates hundreds of millions of dollars in annual savings directly related to improved FM operations around the world.

"We have really taken the value proposition of our organization to a whole new level. People wonder how we are able to achieve such dramatic results, but it has really been an ongoing evolution," says Glenn Hodge, former strategic initiative manager for Intel Corporate Services.

CS has proven the power of using predictive space models and objective benchmarking to drive best practices at sites around the world. Intel senior management supports these initiatives at high levels, allowing the models to proliferate and increase in value. The organization's predictive methodology is so functional that CS has recently been asked to apply the tools to Intel's strategic long range facility planning.

Evolving Tools

According to Hodge, the evolution of the CSM can be broken down into three phases: reactive, proactive, and, the current phase, predictive.

"Back in 1998 we were primarily reacting to cost pressures. We experienced a cyclic downturn and our customers were demanding that we cut costs but maintain our reliability and, in some cases, increase services," says Hodge.

In reaction to these pressures, Intel CS implemented a program of internal benchmarking and cost tracking in order to improve service across sites in North America. The reactive phase helped generate noticeable cost savings, but the system for gathering information—based on the exchange of relatively cumbersome Excel spreadsheets—was highly user-dependent so data accuracy was questionable.

During the subsequent proactive phase, Intel CS adopted a more objective approach to facility planning. Using a Web-based benchmarking system called RAT.com (Resource Allocation Tool), and an array of other predictive cost tools, the organization shifted its paradigm from one of reaction to one of forward projection. As the tools became standardized the organization began deploying them at sites in Europe and Asia.

"Once we built the Web system, our benchmarking process became global. The data was much more structured in how it was entered, so the output was considerably more useful. Then we could go to sites and look at what they were doing and know what the real targets were. When we applied this on a global scale, we saw hundreds of millions of dollars in savings annually," says Hodge.

The proactive approach generates proven data that Intel CS can use for accurate benchmarking, providing a "best-in-class" view of how a site should be operating in comparison to Intel's other facilities globally.

"RAT.com has enabled anyone to quickly benchmark themselves to their peers at very detailed levels, using a variety of parameters," says Hodge.

Additionally, there is a considerably higher level of information sharing between regions and sites. Intel's facility design and management efforts are now much more focused.

"We moved from working on a site-to-site basis, to benchmarking the majority of our budget globally, which drives proliferation of the best methodologies across our entire customer base," says Hodge.

In addition to guiding facility management practices, these tools are used by senior management to forecast spending. By comparing budget requests to objective metrics they can make educated decisions about how to distribute funding.

"The tools facilitate a very high level of global engagement. When we produce these models, there's a lot of interaction very quickly," says Hodge.

Modeling the Future

The CSM's current predictive phase began when senior management asked CS to spearhead development of strategic long-range facility planning.

"When we were asked to run strategic long-range planning we started building these projection models into the plan instead of pulling numbers out of the air or just doing really big estimates," says Hodge.

This type of long-term forecasting helps the organization establish its current direction. According to Hodge, the predictive paradigm looks four to five years down the road at where the company wants to be and how to get there.

"Predictive modeling steers business practices to get you where you want to be, as opposed to a more reactive response, where actions are driven by the organization's business partners. Using these models we can look down the road five years at expected growth figures and see exactly what our costs should be," he says.

Cost of Doing Business

Another predictive model currently in development is "total cost of ownership," which is used to predict the total costs incurred by a facility over the course of its entire lifespan. This formula uses established affordability targets to challenge new facility design.

"The total cost of ownership model is a way of quantifying all of the costs associated with a facility throughout its life cycle. It allows our construction side and our technology development side to look at the real picture of what a facility is going to cost overall, not just offering a price tag for how much it is to build," says Hodge.

Because Intel capitalizes on the use of redundant design features in new facilities to maximize efficiency, considering the long-term impact of decisions becomes significant.

"We're looking at the true costs of operating a facility. If we install a few extra chilled loops in a building where the expenditure is relatively small, it becomes standard and gets copied across our network. Multiply that figure by 11 factories and all of a sudden that cost is substantially more. Then consider the cost to operate three chilled loops versus what, theoretically, could be done with one, multiply that again by 11 factories, and it really adds up," says Hodge.

Key Lessons

Using fiscal space-use models to analyze challenges and identify business opportunities helps CS increase operational efficiency and competitiveness across all of Intel's sites. In the course of eight years, the organization has learned a number of important things about this form of predictive modeling and benchmarking.

Reactive responses are short-lived and often result in departmental fragmentation and defending of fiscal territory. Proactive activities are less agitating and result in more focus on common learning, but predictive models have the potential to be the most powerful because they drive business activities from the ground up.

One important factor is that, according to Hodge, the more sophisticated the modeling effort, the more senior management support is needed. CS has repeatedly proven the value of these tools and, consequently, the organization receives a high level of senior management support for its modeling-based initiatives.

"Without core senior management support and a central operations platform it is very hard to get the globalization effect," says Hodge.

Another key lesson is that, in addition to having high level management support, sustained strategic affordability programs must be adequately planned for and invested in.

"The only way to be significantly different five years down the road is to start planning for change last year. If you don't approach it like that, then five years from now things are going to look much the same as they do today because opportunities for change will have been lost," says Hodge.

By Johnathon Allen

We welcome your Questions and Comments

Copyright 2008 Tradeline Inc.
All Rights Reserved
ISSN: 1096-4894