A Business Executive's Guide to Modern OR
Survey of opportunities to boost profit, reduce cost and increase productivity
By Randy Robinson
What opportunities to improve a company's performance, perhaps dramatically, still wait to be discovered by the alert business executive? I mean notable improvements like higher revenue, lower cost, increased market share, reduced risk, greater productivity, faster turnaround and more efficient use of limited resources.
The high-tech field of operations research offers numerous excellent opportunities to boost performance immediately. Yet time remains to skim the cream before everyone wakes up to the prospects. When most do wake up, companies that aren't taking full advantage of OR will leave serious money on the table and be outflanked by competitors.
For example, here's what former CEO Bob Crandall of AMR Corp. (parent of American Airlines) said about one result of applying modern operations research:
"I believe that yield management [a method of pricing and booking seats, created by AA's OR group] is the single most important technical development in transportation management since we entered the era of airline deregulation in 1979. ... The development of yield management was a key to American Airline's survival in the post-deregulation environment." 
In another example, Nick Donofrio, senior vice president, IBM Technology and Manufacturing, said that the "Asset Management Tool [for supply chain management, from an OR team in IBM] ... helps our business units identify the 'optimal' supply chain strategies. ...[It] allows for enormous cost-saving, inventory reduction, customer responsiveness, and speed of delivery benefits ... [and] has totally transformed our global operations." 
After today's headline frontiers such as e-commerce become routine, operations research may well lead the next generation of vital business facilitators. To discover why, let's tour the field of modern OR, from the executive perspective.
We'll focus especially on fundamental questions: What are the bottom-line benefits? What exactly is OR? If it's really helpful, why isn't it already widely known? What methods does it employ (in plain language)? Where are the better applications? How would an organization that isn't already benefiting get started?
What Benefits Should Companies Expect?
Reflecting on the next stage of e-commerce, Andy Grove, CEO of Intel Corp., offered the following remark in the Wall Street Journal: "Think of every enterprise, whether it is Intel or a hotel, involved in adjusted capacity-demand pricing. ... If this can be done computers-to-computers in real time, you'll see another power of 10 increase in the efficiency of the work in the economic system. Now how do we get there? ... Bits and pieces of it exist in decision theory and operations research." 
The OR ideal is to turn around the declining enterprise, save the project in trouble, and win the war. While typical achievements are more modest, at its best, OR delivers blockbuster results.
What are the specific benefits? They vary widely. Generally speaking, OR helps improve effectiveness. Here are some illustrative sources of effectiveness that is, particular benefits delivered in OR projects:
These broad categories overlap, and more than one might apply in a project.
OR? What's That?
One reason why OR presents untapped opportunities is that the field frequently is invisible or, if visible, misunderstood. Modern OR can be defined as the application of scientific methods to improve the effectiveness of operations, decisions and management. By means such as analyzing data, creating mathematical models and proposing innovative approaches, OR professionals develop scientifically based information that gives insight and guides decision-making. They also develop related software, systems, services and products.
In other words, OR professionals specialize in doing advanced analysis to create and interpret valuable knowledge. The purpose of that knowledge is to assist management by furnishing insight and guidance. OR professionals, first and foremost, are experts in mathematical modeling for managerial applications.
OR may assist with almost any management function. For instance, it helps executives and managers make key decisions, solve urgent problems, design better multi-step operations (often called processes), set policy, plan, forecast and measure results. OR serves non-managers as well. For example, it may help engineers or consumers improve decision-making and other functions ordinarily performed by managers. You will find applications of OR in diverse types of organizations, at diverse organizational levels.
OR professionals also develop software systems, including the large decision-support systems that perform make-or-break jobs in some companies. And they develop new products and services that depend on OR methods.
To sum it up, OR helps take advantage of the latest analytical and computer-system methods to improve bottom-line results. It provides the highest-tech management consulting on the planet. It develops decision-support systems and managerial command-and-control systems, as well as modern products and services that embody advanced decision-making methods. The analysis OR can deliver is vastly more powerful than non-OR analysis such as projections of accounting reports and other standard spreadsheet studies. The beneficial results of OR analysis may be substantial company-rescuing in some instances.
If OR Is So Far-Reaching and Valuable, How Has It Avoided The Spotlight?
Despite 60 years of solid accomplishment, the operations research world remains largely unheralded. Several reasons explain this. OR operates ordinarily behind the scenes. And its professionals tend to be doers, not promoters. In many cases they have shied away from publicity. Furthermore, many older-generation executives, the early prospective customers for OR, were content with far less than cutting-edge analysis to support decisions. Some older executives also felt an aversion to computers and mathematics.
Another significant reason for the OR community's low profile is that they never completely "branded" their field. Consequently, operations research is practiced under many names. These alternative names include the near-synonyms management science, decision technology, decision support, policy science, system analysis (where the applications are related to management and decisions), management technology and management analytics. Other aliases, adding to the confusion, apply to special subjects. Examples are operations analysis in manufacturing and the military, financial engineering in finance, marketing engineering in marketing, rocket science in Wall Street, and a recent arrival, focused on a few of the analytical methods, complexity science.
As if that weren't enough, yet another source of confusion is that most traditional branches of engineering and of applied mathematics practice OR within their own fields. For instance, OR methods are employed and OR applications are worked on by aeronautical, chemical, civil, electrical, industrial and mechanical engineers (to name a few), and by statisticians, actuaries and other applied mathematicians. The same is true of other applied scientists, notably economists and their close colleagues, the corporate finance specialists.
Despite such an embarrassment of riches in names, and despite being practiced often by those who prefer the banner of another field, OR is OR is OR. The name operations research, while misleading (neither is it research in the popular sense, nor is it limited to the subset of activities called operations), is far and away the best established name.
On the bright side, the fact that OR is practiced under all those different names and within all those different professions underscores the point that operations research is a wide-ranging, influential subject with staying power. It's no passing fad.
Large Number of High-Tech Methods
An OR professional, like a medical doctor, is expected to select from the total array of the latest technology whatever promises to do the most good in the task at hand. This technology includes methods that might be found in any branch of mathematics, any of the sciences including the social sciences, and any branch of engineering.
Over the years, members of the OR community have invented important new methods and significantly advanced existing methods. But the OR professional should select among all helpful techniques, not just those created within the OR family.
Because of the breadth and depth of the growing collection of pertinent methods, combined with a wide range of application fields, OR, like medicine, is divided into specialties. The generalists may call in specialists and work in teams.
When first encountered, methods commonly used in OR seem obscure. This impression is fostered by technical labels such as multicriteria decision analysis, linear and nonlinear programming (nothing to do with computer programming), discrete-event simulation, queueing and other stochastic process modeling, conjoint analysis and neural networks.
We can clarify OR techniques despite their labels, however, because most projects apply one of three broad groups of methods:
Where Are the Applications?
Because operations research can make and has made contributions in virtually all industries, in almost all managerial and decision-making functions, and at most organizational levels, the list of past and prospective applications is prodigious. To sample from the application universe, let's briefly look at OR in major business functions (finance, production and so on), and then consider some applications that cover multiple functions (such as revenue management and supply chain management).
This sampling of applications won't try to go industry by industry. All industries contain many good opportunities to benefit, including industries that might at first seem unlikely clients e.g., agriculture, health care, broadcasting, entertainment and sports.
Here are selected business functions, with a few comments about OR applications in each:
Board room and senior executive offices. Top-level, high-impact policy-setting and decision-making offer companies especially great potential to gain competitive advantage. The power of OR to reveal the implications of different courses of action, and to find a preferred course of action among numerous possibilities, goes far beyond that of ordinary spreadsheet analysis.
Senior-executive applications include, for example, mergers and acquisitions, major expansions, divestitures, valuation of companies or operating units, facility location, and redesign of the entire supply chain. In companies that operate through large contracts, the ability of OR methods to help them with price setting and risk reduction in competitive bidding are keys to gaining an edge.
OR techniques make it possible to quantify risk, always vital information but seldom quantified in executive offices. OR methods also can quantify and balance other qualitative aspects normally thought to be beyond the reach of analysis. In addition, some OR methods help with gathering and displaying inputs from different executive participants, and with building buy-in to the final decision.
Many excellent applications come to the top-executive level. Even a slight percentage improvement beyond what the ordinary analysis discloses will produce a significant payoff for the company. Often the improvement is so substantial that it makes a critical difference.
Manufacturing and service operations, distribution, transportation, telecommunications. The "operations" in operations research suggests that this application category is the original home of OR. That's not quite correct because the name comes from military operations, not business operations. Nevertheless it's true that OR successes in contemporary business pervade manufacturing and service operations, logistics, distribution, transportation and telecommunications.
The myriad applications include scheduling of all kinds, routing, improvement of work flow and elimination of bottlenecks, inventory control, business process reengineering, site selection, greenfield facility planning and general operational planning. Two growing applications that are distinguished by their use of several methods to cover several functions revenue management and supply chain management (both discussed below) are in this realm.
Manufacturing companies make heavy use of simulation. Transportation and telecommunication companies make heavy use of optimization. In fact, most OR methods are well exercised throughout this arena.
The names operations analysis and operations management often appear as synonyms for OR in operations.
Finance and control. The core financial applications occur in investment, lending and borrowing, financial planning, and insurance. Financial institutions have these, as well as applications in their operations. In a manufacturing or service company, the applications are those within the domain of the chief financial officer.
Mathematical models, signatures of OR, have become common in finance. Much publicity surrounds the investment applications, especially in connection with newer investment media like financial derivatives. The index fund was invented by OR people. So were key advances in investment performance measurement. But past and prospective OR successes cover a much wider range, including company financial projections and forecasts, valuation of companies and product lines, selection of corporate borrowing vehicles, hedging, foreign currency analysis, treasury operations, analysis of taxes, asset and liability management in commercial banks, and credit approval.
The name financial engineering is popular these days. That name typically serves as a synonym for OR in finance. The Wall Street version of the name rocket science usually means OR in investment.
Marketing. As in finance, the opportunities to gain a competitive edge through quantitative analysis are multiplying in marketing. Those opportunities arise, first, in strategic marketing. Strategic applications include general planning, sales forecasting, market selection, market measurement, segmentation and targeting, and product portfolio analysis. Other opportunities present themselves in tactical marketing. They include decisions about pricing and promotion, salesforce sizing and deployment, location of distribution facilities, advertising and product design.
You'll find these applications in manufacturing and service companies. Comparable applications are found in pure marketing firms, including ad agencies and market research companies.
Paralleling the name financial engineering is the currently popular name "marketing engineering," which ordinarily means OR in marketing.
Information technology. The impact of OR on IT is growing. The top IT job in most companies soon may require a strong background in OR.
For many years computer vendor executives have been saying that the real payoff from computerization is the ability of computers to produce crucial management information. OR is the field dedicated to creating the content of management information. OR brings the best methods to bear on deciding what input data to collect, how to convert raw input data into valuable, decision-guiding information, and how to present that information for greatest managerial usefulness.
OR is behind a growing number of IT buzzwords, including decision support systems (DSS), online analytical processing (OLAP), and optimization components (along with other analytical components) of enterprise resource planning systems (ERP) and of e-commerce systems. Generally, OR furnishes the expertise needed to develop management-oriented software and communication systems with analytical content, such as shop floor planning and scheduling systems, data mining systems, intelligent pricing (revenue management) systems, supply chain management systems, and real-time command and control systems.
Because OR work is so closely connected to computers and data communication, OR and IT professionals often become partners in projects.
Human resources. The military services apply OR extensively for personnel planning and related HR purposes. While not as many HR applications in business have been made known outside client organizations, the opportunities, where choices are made in complex situations, are increasing. The most likely topics include analysis of retirement plans and other benefit plans, compensation studies and compensation planning, compliance with regulations, and reduction of travel expenses.
Legal. A classic OR application is deciding how best to proceed in a large-dollar legal action. The OR method called decision analysis is perfect for sorting out and correctly weighing the many uncertainties and contingencies in this situation.
OR professionals also may provide the expert analysis of quantitative data sometimes needed to prepare for court cases.
Engineering. Such OR techniques as decision analysis, the analytic hierarchy process and mathematical programming can assist with engineering decisions for example, in making choices in design and in testing.
The engineering professions make heavy use of linear and nonlinear programming developed by the OR community. For example, these optimization methods are commonly employed to help manage oil refineries.
Research and development. The classic R&D application is selecting the research portfolio that is, deciding how much to invest in each of the possible R&D projects.
Revenue management. Turning now to a different view of applications, let's consider some that involve multiple functions (and often multiple OR methods). We'll begin with a current hot topic: revenue management.
Also called yield management (praised by former CEO Bob Crandall of AMR, above), revenue management entails accurately forecasting demand and then adjusting prices over time to more profitably allocate fixed capacity. You might term this super intelligent pricing. It shares much in common with the application of competitive bidding. Developed originally for pricing and booking of airline seats, the approach has spread to hotels, rental car firms, broadcasting, and more recently to manufacturing.
Supply chain management. Supply chain decisions dictate the who-what-when-where all the way from purchasing and transporting materials and parts, through manufacturing products, and finally distributing and delivering products to customers. The prime management goal might be to reduce overall cost, very large in many companies, while filling customer orders even faster than before.
The power of OR methods makes it possible to examine this complex chain comprehensively and to search among a huge number of combinations for those that seem most beneficial.
Data mining. As interest in the analysis of customers and buying behavior grew, along with the ability to accumulate data in extremely large databases, data mining became a major force. OR methods give the means to find in any large collection of data (not just customer data) those patterns, trends and connections that will assist management.
Complex scheduling. OR brings the latest techniques, and decades of experience, to the scheduling of resources in complex situations. Applications abound, especially in manufacturing, service and transportation industries.
Intelligent real-time systems. Modern decision-support systems respond in real time, and draw analytical capabilities from a combination of mathematical and AI/expert system methods. Already important in the military, these OR-developed systems are likely to become important in business too.
Making Effective Use of Modern OR
Perhaps a company wants to try OR for the first time. Or perhaps it wants to strengthen an OR activity in place. How should it proceed?
Any exploration in OR inevitably surfaces the question: How do you recognize qualified help? Look first at the usual dimensions working track record, personal qualities and academic training. What may assist you further is to be aware that the OR field, through collective experience, has arrived at unwritten but widely known practice guidelines the professionals follow. My interpretation of those guidelines, in a nutshell, is:
Another question is whether to seek OR expertise outside, or to staff an OR function inside, or to utilize a combination. You know what fits best within your own organization. I'll just make this comment. Successful OR, in my experience, takes teamwork between the client and the OR professional or professionals, usually with company members close to the client participating. Great OR success almost always comes from great teamwork.
A third question arises in the event you want to staff OR inside the company: Where to position it? If you form a single OR group for your company, I believe they should go where they can report to an OR champion in senior management, and where they can serve clients unobstructed by chain-of-command problems. In my view, most companies eventually will discover that OR is sufficiently valuable to merit distributing it to the various units and departments as IT, accounting and marketing are distributed, for example. The independent OR specialists will do better if they have a dotted-line corporate homebase comparable to corporate homebases for IT, accounting and marketing.
To the foregoing, I add two bits of thumbnail advice to prospective clients who want OR to help them arrive at better decisions or policies. Never follow an analytically based recommendation that doesn't feel right. Be aware that the analytical backup you get will not in every case produce a radically original, counterintuitive answer; sometimes it serves to confirm what seemed best at the outset (but lacked analytical support).
If you decide that you want an impartial referral to outside assistance, try the referral service of the nonprofit Institute for Operations Research and the Management Sciences (INFORMS), the OR professional society which publishes this magazine. Call toll free 1-800-446-3676, visit the Web site www.informs.org/Executives, or send e-mail to firstname.lastname@example.org.
For further reading see the sidebar Billions in Business Benefits.
The author expresses appreciation for contributions to this article from long-time friend Carl Harris, a pillar of the OR community, who passed away suddenly in April. The author also thanks colleagues on the Public Information Committee for their suggestions: Lew Pringle, Newt Garber, Gary Lilien, Dick Larson, Saul Gass, Sid Hess, Burnell Brown, Howard Finkelberg, Reginald Joules, Mark Doherty and Barry List, and OR/MS Today editor Peter Horner.
Randy Robinson (email@example.com) is working on a book for management about OR. A career OR practitioner, he chaired the ORSA-TIMS board that organized the creation of INFORMS and he served as founding INFORMS executive director.
OR/MS Today copyright © 2000 by the Institute for Operations Research and the Management Sciences. All rights reserved.
Lionheart Publishing, Inc.
506 Roswell Street, Suite 220, Marietta, GA 30060, USA
Phone: 770-431-0867 | Fax: 770-432-6969
Web Site © Copyright 1999, 2000 by Lionheart Publishing, Inc. All rights reserved.