Getting Your Act Together: by Policy

Brendan Collins and Ernest Huge

eeking dramatic business improvements in MBP was originally developed by Japanese customer service, cost reduction, and safety, companies as a defensive capability for dealing Sthe Southern Pacific Lines (SP) used Man­ with large unforeseen events. The pioneering agement by Policy (MBP) from the beginning of companies were often victims of soaring oil costs their Total Quality (TQ) journey. SP utilized the or dramatic loss of market share. However, MBP support of a core group of seasoned quality practi­ has evolved into a major offensive weapon. MBP tioners from the inside and hired from the outside. provides the focus, the means, and the driving They began after only 2days of Total Quality edu­ force for excellent companies that aspire to domi­ cation for the management group. Basic process nate world-wide markets. improvement tools (primarily the Pareto chart and Although cynics sometimes state that MBP cause and effect diagram) were introduced and is merely the way Management by Objectives systematically used during regular reviews of (MBO) is supposed to work, there are several key progress against the objectives. differences. MBO is frequently limited to the SP's approach to MBP was to "just do it," financial objectives of the company; MBP focus­ learn by doing, use TQ methodologies, and contin­ es on a few major corporate operating perfor­ uously improve the MBP process. They reasoned mance improvements or opportunities - the that applying selected TQ principles to a vital few strategic "hows" behind financial results. MBO objectives, a major MBP tenet, was the better thing is geared almost exclUSively toward company to do. Also, they felt that this approach would inte­ needs; MBP focuses on customer needs. It mar­ grate TQ with management of the business from shals the means within the company to satisfy the start. Their approach has been successful ­ those needs. MBO focuses solely on results. MBP SP has met or exceeded targets during 1992 and focuses on both results and the processes by 1993 in many key focus areas. which results will be achieved. What is Management by Policy? MBO translates company directly MBP is a process. It into functional and individual performance is also called policy management, policy deploy­ objectives; MBP is more comprehensive. MBP ment, hoshin planning and, in Japan, hoshin concentrates on the improvement of processes kanri. The term hoshin planning is limiting that cross functional lines and departments, so because the process involves both planning and top management leads through process instead execution. managing only for results.

20 Target Volume 9, Number 6 Factors in Totallmllrovement Process

SUCCIlSS Factors MBP is for companies ready to manage by facts and willing to engage in a rigorous interactive process to select the right objectives for major corporate improve­ ment and to align everyone behind them. The objectives developed through the MBP process should: • Represent major opportunity for improvement from Figure 1. Summary of the various factors which must be considered in order to select the vital few objectives fora total the customer's perspective. organization's improvement processes. • Be crosscfunctional. • Deserve high-level management attention. MBP is obsessed with the customer. All improve­ • Be urgent, such as when a significant threat is ment activities are driven by customer needs, focusing encountered. With the Japanese, the oil shortage in on a few major improvements or opportunities. It gives 1973 and extreme currency fluctuations are examples. careful and det~iled attention to process, means, and Both provided astrong impetus to develop MBP. results. It recognizes that most major improvements The Three Phases of MIl' will require cross-functional cooperation. To identify Each of the three phases of the Management by major improvement objectives and the means to Policy process must be successfully executed to have a achieve them, MBP requires a high degree of participa­ successful process. The three phases are: policy establish­ tion by the people who must realize the goals. ment, policy deployment, and policy implementation. The participatory nature of the process must be Phase I: Policy Establishment emphasized to keep it from degenerating into another The first phase is a process of selecting the right exercise. Many companies can use the tools of quality objectives. It considers the customer's needs, the busi­ and have a zeal to serve the customer, but their efforts ness environment, the company's long-term plan, "chase too many rabbits." To focus their effort, they internal issues and capabilities, current performance, need a strategic methodology that incorporates honest critical success factors, and regulatory requirements. reviews of what is actually being done -'- and what can Customer satisfaction is the central driving force. be done. That's MBP. That is a fundamental premise. Consequently, the pro­ Most U.S. companies that have set out to imple­ cess starts with identifying customer needs and deter­ ment MBP - ALCOA, AT&T, Blount, Digital Equipment mining customer satisfaction. Actions to determine Corporation (DEC), DuPont, Florida Power &Light, these needs vary from, "We already know our customers' Hewlett-Packard (HP), Intel, Proctor &Gamble, South- needs, so let's stop wasting time," to "Oh no! We have . em Pacific Lines, Texas Instruments, some units of Ford to come up with a million dollars to get the best survey and GM, and some automotive suppliers - feel that it people in the country on board." has significant advantages although they may be still There is a happy medium. While it is important to short of world class in the implementation of MBP. ask the customers what they need, it is also important Who's world class in executing MBP? Canon, Toy­ to first figure out who the customers are. The complaint ota, Aisin Seki, and a number of Japanese companies records will identify the complaining customers and that have won the Deming Prize during the past decade. some ·of their needs. But don't forget about those who So far as is known, the American companies engaging in do not make an effort to complain. MBp, by whatever name, have reached the same levels of Policies are formulated by reflecting upon the proficiency, although at the time of winning the Deming company's current status. (See Figure 1.) Companies Prize, Florida Power &Light was judged to have a pro­ accomplish this reflection differently. For example, the cess comparable to many leading Japanese companies. president of Komatsu goes alone to a mountain retreat 21 Target Volume 9, Number 6 Catchball

Once the critical areas for improvement have been selected it is important to determine how progress will be measured. Measurements are called indicators, which suggests not confusing a measurement with genuine accomplishment. An indicator of an important output of aprocess is called a point, whereas indicators for process parameters are called check pOints. Long-term and intermediate "draft targets" must be proposed for each indicator. An improvement target communicates the degree of change reqUired. Targets frequently cause discomfort. Those who work with a process may be reluctant to be pinned down to a con­ crete number denoting a level of improvement that must be pursued. They may simply not want to be held accountable, or they may believe that the MBP process has not yet resulted in an analysis thorough enough to substantiate the desired . The draft target is the result of the best analysis available from the corporate perspective. In alllikeli­ Flgu,,2. hood it does not have the benefit of local analysis or identify the capability of local processes. Furthermore, for several weeh armed with voluminous data. He up and down the organizational ladder those who are returns with suggested policies for the next year. In some expected to bring about the necessary changes must companies, the top management team collectively "buy-in" to each target. debate until they reach consensus on the significant few. For example, a policy might be the reduction of In others, suggests the initial poli­ emergency field service calls. The indicator would be a cies. In some, both top and middle management reflect specific definition of emergency calls and a formula for in parallel. Asking yourself the following questions will measurement, along with the current performance help ensure the quality of the strategic analysis: number. The draft target might suggest reducing the • How will you measure superior value in the future? number by half within a year. • How do you know what it takes to win? The policies, indicators, and draft targets should • What are the sources of competitive advantage? Of be published company-wide well in advance of the bud­ market presence? Of market access? Of core geting process so that the areas singled out for dramatic competencies? Of work processes? Of organizational improvement are highly visible to everyone. Wide visi­ structure? bility allows those most affected to complete the analysis • What gaps must be bridged? What do you need to be and identify the improvement actions they must imple­ able to do that you're not now doing? What can you ment to achieve each draft target. stop doing? What processes must be improved? Phase 1/: Policy Deployment • What is the motivating and aligning goal of the Phase two, policy deployment, is making sure that organization? all of the right people are included and that they have a Policies should then be prioritized by considering clear picture of what they must do. Contributors must several factors, principally: importance to the customer, have confidence that they are working on the right opportunity for competitive advantage, and areas with things at their level. Their targets need to be both chal­ the greatest need for improvement (the worst perfor­ lenging and achievable. mance relative to the competition). Ultimately, the goal The scientific method (Plan, do, check, act ­ is to identify the critical few things to work on. PDCA or the eqUivalent) is used to conduct analysis. 22 Target Volume 9, Number 6 Deployment of an Objective CORPORATE Top managemenVcross-functional committee for deliveries objective/target is : • Reduce customer order leadtime by 50%.

Then consistency of direction is ensured, and clear link­ age from top to bottom is established. The top-level objective merely points toward an opportunity. To get to the next level the question, "Why is this a problem or opportunity?" must be answered. This process of ques­ DEPLOYMENT tioning is followed down through the organization to the • Order entry by 50% furthest point possible. Analysis should answer the ques­ • Sales by 30% tion, "Why?" This analysis becomes the thread that con­ • Manufacturing by 20%. nects the strands of deployment into a single, tightly­ woven fabric. Good analysis requires that answers be sup­ ported by facts and data. Participation and buy-in is assured through "catchball." Catchball is an iterative process: developing objectives and the plans to obtain them, sharing them with persons who must execute the plans, requesting and DEPARTMENTAL OBJECTIVE/TARGET PROJECT OBJECTIVES considering their input. After sufficient involvement and • Order Entry Reduce by 50% 1) Implement an on-line order commitment by all affected parties, objectives and plans acceptance and control system are finalized. Catchball is two-way communication that by 7/15 results in joint commitment and joint ownership. (See • Sales Reduce by 30% 1) Improve the order form by 6/1 Figure 2.) 2) Provide additional training by 8/1 Figure 3 shows how objectives and means are cas­ • Manufacturing Reduce by 20% 1) Implement product focused cells caded throughout the organization. Indicators are estab­ for product ABC by 9/30 2) Reduce set-up times to ten lished for objectives and means respectively. In this minutes or less by 8/31 example, the policy is to reduce the customer order lead­ 3) Increase machine availability to time of a product that is built to order. 95% by having operators perform routine maintenance by 12/31 Phase 11/: Policy Implementation 1) Redesign product ABC so that The third phase, policy implementation, is possible • Engineering customer orders can be custo­ once the areas for improvement are selected, the appro­ mized in final assembly by 7/30 priate people are on board, the analysis has been done, • Human 1) Finalize work rules with the union and the contributions are identified as sufficient to Resources to allow operators to perform achieve the overall corporate target. routine maintenance by 2/15 The MBP system can be visualized as the systematic Figure 3. MBP cascades policy in more detail throughout the organization. spinning of a spider's web shown in Figure 4. Its structure is intricate and delicate, so it can be easily broken. With­ THE MBP Spider Web out close care and attention it can become an entangled CEO mess. It does not appear spontaneously because it is the product of skill and intensive effort. The "spider's web" Managing Objective leader has several anchors around the outside, so that each seg­ the details ment makes its own unique contribution, but the full impact is the product of all the pieces. Then the whole is greater than the sum of the parts. Progress The CEO is responsible for ensuring that the sys­ CEO's reviews management tems are in place to guarantee the quality of products or team services, timely delivery, reasonable pricing, safe prod­ Cross-functional ucts, and community confidence. Each of these is management required to ensure that a proper balance is provided in Figure 4. 23 Target Volume 9, Number 6 Progress Reviews Level Frequency Reviewer Review.. I 3/Year Senior Executive Department Head II 4/Year Department Head Middle Managers III Monthly Middle Manager Supervisor/Other CEO Monthly CEO/Other Objective Leaders and Highly visible, and not leaving good intentions to Executives other presenters chance, the objective leader is the conscience of the cor­ poration straining to achieve dramatic improvement. Flguls5. Effective participation by the CEO's management team is imperative. The choice of the word "team" is the pursuit of customer satisfaction and the quest to significant. In business, sports, and entertainment, we remain profitable. live in an age of superstars, but when success is The CEO's support team consists of the heads of the achieved, teamwork rises above personal idiosyncrasies. line organizations, and their roles also cross functional Teamwork behavior depends on having a common lines. They help the CEO ensure an optimum balance in vision, practicing interdependence, understanding the meeting customer needs. In larger companies it is relationship between personal success and group suc­ impossible for the CEO to pay the kind of attention to cess, focusing on the common mission, and skillfully detail that is necessary to set and achieve targets. reconciling conflicts. The Ob/ectlve Leader Personal styles of management are submerged in Each improvement objective selected should be an MBP process. Only by working with a team engaged given a "home." Therefore each objective leader (in Fig­ in an MBP process does the CEO establish or confirm the ure 4) is responsible for leading the effort to achieve a corporate mission and vision. The team develops com­ particular objective. Objective leaders are generally pany strategies based on objective analysis and assess­ senior executives with a thorough working knowledge of ment. They identify critical objectives for corporate the functional areas spanned by the objective. improvement. They monitor progress achieving them. The objective leader conducts an analysis to deter­ They allocate and manage resources consistent with the mine current performance and quantify it. Considering mission, vision, and major improvement objectives. the capability of existing processes to support desired They also constantly check to see that the infrastructure performance, the objective leader determines where the facilitates clear communication of their intentions and performance must be in the short-term and in the mid­ mirrors their behavior. term. Since the objective leader must cross functional Cross-Functional Management lines, this person must have the authority to cut through "Cross-functional management" well describes the bureaucracy which is all too often intent upon pro­ how the process actually works. If literally managing tecting its turf. across departmental boundaries sounds revolutionary, it's For example, reducing the time from design con­ because organizations tend to be territorial in nature. It's cept to market is often an important objective. In some natural to refer to "my" department, "my" people, "my" Japanese companies it is so important that the president equipment, or "my" product, but in most organizations is the objective leader. Clearly, this objective or policy is this inhibits seeing oneself or one's co-workers as contrib­ cross-functional. The leader and team must understand utors to an overall process. Although it becomes critical the current process and identify its strengths, weaknesses, further up in an organization, cross-functional manage­ and vital few opportunities to achieve the market needs. ment is important at every level. Usually the objective leader and objective team Progress reviews are another critical element in the commission ad hoc teams to work on a few significant management web. Reviews should be constructive rather processes (such as concurrent design). By acting as liai­ than destructive. In a destructive review the reviewer son with upper management, the objective leader is also vents frustration, plays out personal agendas, exetcises the link to the all-important cross-functional aspect of inappropriate power of position, or otherwise demeans management at the top of the company. the person being reviewed. In addition, a destructive This person is the catalyst who ensures that the review focuses only on results. necessary steps are taken on time, and that progress is Aconstructive review, on the other hand, focuses adequate. An objective leader is a coach who encour­ on both results and the process for obtaining results. By ages, lobbies for needed resources, teaches, gives feed­ using data describing both processes and results, a back, and makes sure that the schedule is maintained. reviewer follows a path of diagnosis to more accurately 24 Target Volume 9, Number 6 determine the current status of progress and identify the capability of TQ implementation, in addition to gaps and weaknesses. In addition, the visibility of such traditional strategic planning inputs. data creates a basis for action that can be tied to a • Focusing strategic initiatives to a greater degree than schedule and that can be better understood by everyone. before. Aconstructive reviewer feels jointly responsible for both • Deploying high level goals in a more consistent way results and for the means by which they can be attained. throughout the organization. Figure 5 is a gUide to ensure a systematic • Implementing an ongoing process to assess TQ approach in conducting reviews. Unless systems elimi­ implementation, (usually using the Malcolm nate the need to solve the same problem again and Baldrige National Quality Award criteria). again, and are used with meticulous follow through, MBP becomes another bureaucratic exercise instead of a One output of the MBP design process is an MBP rigorous driver of improvement. manual, a procedural cookbook for all participants in the MBP process defining who does what by when, why, ApproachBS to Impl_entatlon and how. (Yes, one of those.) By documenting the pro­ The so-called classical way to implement MBP is cess, a manual takes MBP from merely a conceptual the following sequence: First, implement what some peo­ level to an operational level. Equally important, the doc­ ple call the Daily . More specifically, umentation pegs where the MBP process itself is so that you should have a critical mass of competency through­ it can be improved. Actively using a manual to improve out the organization using the PDCA cycle to improve a basic process converts it into a means of joint improve­ processes and the SDCA (study, do, check, act) cycle to ment rather than follOWing a distant prescription by maintain and standardize improvements. Get the major­ someone else. However, care must be taken to ensure ity of critical processes in control and capable. Have at that the manual is at the right level of detail. Trying to least 50-60 percent of all persons involved in formal pro­ construct an overly detailed manual before beginning is cess improvement teams. Second, proactively assess cus­ the classic case of over-planning a new idea before really tomers needs and have a reliable method for translating understanding it. needs into product and/or service reqUirements. Develop Many U.S. companies used the classical sequence aprevention-based approach for critical cross-functional of implementation because they were well along on the business systems. Third, implement MBP. TQ journey before they heard much about MBP. Since Many U.S. companies implementing MBP little has been written about MBP, companies needed to generally followed the classical sequence. Some started do some digging. Hp, one of the first U.S. companies to implementation in the same way that many Japanese initiate MBP implementation, learned about MBP from companies did - evolVing MBO into MBP. In fact, most their Japanese subsidiary, Yokogawa HP, which won the companies started by deploying top management's goals Deming Prize. While HP initi,ally deployed objectives vertically throughout the functional organizations with without deploying the means to attain them, their an analysis of. the means - which is essentially how approach has apparently matured to a very good model. Hewlett-Packard began, for example. ADifferent Approach to Implementation ANew Mind Set Is Not Easily Acquired On the other hand, Southern Pacific is representa­ Initially many American companies implement­ tive of a few companies that heard about MBP during the ing MBP were concerned only with goals; means or initial awareness stage of TQ, and decided to use a limit­ methods weren't considered. There was little analysis, ed form of MBP from the earliest stage of implementa­ catchball, or cross-functional coordination. They set far tion. They identified several overarching business objec­ too many policies (such as ten to twelve, when three is tives (such as improve service reliability, improve loco­ as many as most companies can concentrate on). Even motive utilization), and began limited use of TQ so, these companies feel that they received significant methodology to marshal their limited expertise toward benefits from the following initial steps: reaching them. Companies having little experience with • Consideration of external and internal customer TQ analysis deploy the policies as best they can. South­ needs, competitive benchmarking, and the current ern Pacific used regular reviews of progress vis-a-vis 25 Target Volume 9, Number 6 Management by Policy MBP Policy Deployment Stage I Stage II Stage III Stage IV Stage V Vision General qualitative Vivid qualitative Vivid description and Vivid description and Operational definition and some objective goals comprehensive goals comprehensive goals Strategy to Delegated to a Internal process; Developed by top Internal capabilities Obvious competitive attain the planning group; no capabilities not management; fair considered niche supported by vision obvious competitive considered competitor intelligence analysis advantage or niche Annual General General; indicators Focused; indicators Indicators and targets Contribution to policies/ and targets not and targets required required; some consi- improvement based upon objectives to required. Little use but little analysis. deration of means to analysis of means to execute the of basic or planning Some use of tools. obtain objectives. Wide- obtain objectives strategy tools. spread use of tools. Policy Nonexistent Input from one Input from two Input from three Input considered by all deployment management level. management levels. management levels. who must execute. All degree of One level Two levels Three levels levels catch ball Integration None Some integration Moderate - but little Fair degree of Integrated indicators and degree of integration integration across and down deployment organization Reviews- None Occasional at top Regular at top; Regular at some levels Regular at all appropriate frequency level of management occasional at next level but not at all levels levels Review Judgmental of Some interest in Interested in means Focused more on Purpose is to seek cause process people; only means and results process; no interest to and to better understand interested in results place blame processes which cause results Figure 6.

objectives to initiate use of several old quality tools (that • Inadequate planning detail resulting from lack of is Pareto and cause and effect diagrams). Implementing planning involvement by those who must execute the in this way integrates TQ efforts and business objectives, implementation plan. Using no MBP manual or which is a major hurdle for many companies. using it bureaucratically. There is no "one size fits all" implementation pro­ • Failure to use the PDCA cycle on the implementation cess for MBP. The approach needs to be customized to process. the current capabilities of the organization. Figure 6 is a • Not understanding the MBP manual- not learning rough guide to the degree that MBP can be employed "how." depending on the maturity level, or stage, or the organi­ zation. To decide where to begin, conduct an organiza­ • Not managing the expectations of internal customers; that is, creating the expectation of a tional assessment of current capabilities with respect to TQ in general and to the elements of MBP. The assess­ world-class model, but implementing a lot less. For ment need not be elaborate or expensive. example, implementing a version of MBO and Some of the most common pitfalls are: calling it MBP. • Trying to do too much too fast beyond the organi­ • Using destructive review processes in which zation's capability, such as trying to implement at managers continue to blame people for poor results once all elements of the world-class model. One can instead of helping to diagnose and improve processes easily see how this would set up a paperwork that affect results. nightmare - just another bureaucratic process that • Insufficient analysis to determine means/ counter­ ultimately topples from its own weight. measures. • Insufficient preparation of the organization for a • Lack of of focus - inability to settle on only one to major three policies.

26 Target Volume 9, Number 6 • Calling a one-way dictate from top management Summary catchball or two-way interaction. The integrated approach must make sense in the • Delegation of policy development by top manage­ environment in which it is proposed. It is nota planning ment to a staff planning function. model imposed outside the normal business planning, but an integral part of running. the business to structure What Make, Management b, 1'011c, Dlffere"t? challenges and change. MBP requires persistence. Above all, it stretches management skill in walking the fine line between Many companies implementing TQ are having broad perspective and attention to detail. trouble integrating it with business strategy and execu­ The discipline of MBP forces a company to cast a tion. Quality is just another item on an already full plate. Management doesn't use quality principles in their daily cold eye on reality and, from the customer's perspec­ life. There are too many strategic objectives. Management tive, prioritize improvement processes. The ability to can't concentrate on them, let alone the total organiza­ select the vital few is central to the achievement of dra­ matic improvement. tion. What is missing? Astructure to assure that manage­ ment uses quality methodologies to do their job. MBP is There is no substitute for the human develop­ that structure.. In essence, it is the missing link in many ment required for interaction as stratification and TQ implementation processes. analysis are pursued throughout the organization. The tendency to grab qUick fixes or to ramrod expert solu­ What is at stake is the difference between casual improvement and focused, dramatic, and systematic tions is always inviting. It is also divisive and rarely change. This can easily equate to the difference between results in ownership. success and failure. Analysis is at the heart of the process, so success also depends on the proper use of analytical tools. For Brendan Collins, Ph.D., is a total quality control practitioner. As the most part the seven basicQC tools are sufficient, manager ofpolicy deployment, he was a prime architect ofthe but maturing companies also use new tools such as Management by Policy efforts at Florida Power and Light, the first affinity diagrams, interrelation digraphs, and so on. company outsideJapan to win the coveted Deming Prize. In 1991 The targets set must be correct and consistent Collins joined Southern PacifiC Lines as vice president-quality with organizational capabilities. Analysis is also management systems. He is a Malcolm Baldrige National Quality required to determine the root cause of the gap Award examiner. between current performance and the target. Only then Ernest Huge is an expert on world-class manufacturing and management. To his current position at Deltapoint ofBellevue, is it possible to select the actions that ensure that this Washington, he brings more than 25 years of experience in quality gap is closed. improvement,]ff, and managementjJWtiorIs with Generall)ynamiaY, Clear does not happen by accident. It NCR, Clark Equipment, and FMC. Huge was a Malcolm Baldrige requires designated project managers at the executive National Quality Award examiner, served on the AME Board of level. If structured progress is important to the compa­ Directors, andfounded the newsletter which bealme Target He autIxJrrKi ny, it is worthy of executive attention. When a company The Spirit of Manufacturing Excellence and co-authored Total Quality. embarks on MI3P, it is important that the question, "Why?" be asked as frequently as, "What," which ©1993AME~ For infonnatlon on reprints. contact: merely outlines steps and structure. Answering, "Why?" Association for Manufacturing Excellence creates the understanding behind the structure that 380 West Palatine Road. Wheeling, 1L 60090 7081520-3282 promotes permanence and depth of implementation. ~

27 Target Volume 9, Number 6