HOW TO BUILD A BALANCED SCORECARD
by Arthur M. Schneiderman
The balanced scorecard (BSC) has undergone significant change since its widespread popularization in the early 1990s. Although the first balanced scorecard was an integral part of its creators' strategic planning process, its subsequent emulations focused on it as a simple instrument rather than as one element of a total planning system. Consequently, most early adopters just took their myriad of existing non-financial performance measures and force-fitted them to an arbitrary framework that classified scorecard metrics into the prescribed categories of financial, customer, internal, and learning and growth.
I've chronicled elsewhere the resulting common failure modes. Number one on that list was:
"The independent (i.e. non-financial) variables on the scorecard are incorrectly identified as the primary drivers of future stakeholder satisfaction."
Unfortunately this fundamental misapplication of the BSC concept is still all too prevalent.
However, academics, consultants, and practitioners alike have learned much over the last decade. Leading edge BSC proponents recognize that a meaningful scorecard must be viewed as an integral part of an organization's overall management system. But to build on its brand image, "Balanced Scorecard" promoters have used its moniker to provide a name umbrella over its continuously redefined and expanding boundaries. Today, in best-practice organizations, the BSC is tantamount to their business planning system.
But having recognized that the BSC itself is only one part of a comprehensive process, there has still been little documented about that process itself. What has been written describes the method for its creation and use in such general terms that a practitioner is left with insufficient detail on exactly what needs to be done. The objective of this e-paper is to provide my view of that missing level of detail.
In Part 1, I will describe a 9-step process that assures the identification of a manageable and actionable set of BSC metrics that link directly to an organization's strategic objectives. But organizational success - just like a coin or a magnet - has two sides: planning and doing. Successful organizations excel at both. They do the right things and they do them right. My focus in this e-paper will be on the planning side of that fateful coin. I refer you to my other writings on process management (see also my publications) for more on its control, improvement, and reengineering facets.
Part 2 addresses the difficult task of translating strategically chosen stakeholder segment requirements into a prioritized list of internal process improvements. It is the improvement of these targeted processes and sub-processes that will make-or-brake the realization of strategic success. What makes identification of these vital few processes difficult are their many interdependencies and varying impact. Seeing through that cloud of complexity and uncertainty requires the use of some unfamiliar analytical tools and an appropriate balance between established facts and the organization's collective instincts.
Finally, in Part 3, I will describe the fundamentals for extracting the appropriate set of BSC metrics from the near-infinite list of possibilities that still exist even after the vital few processes are identified. Finding those leveraged internal process measures is key in achieving a successful BSC implementation.
*This e-paper was first posted on December 20, 2000 (Part 3 was posted earlier). It will be appearing in hard copy as a Chapter in the Handbook of Performance Measurement, Michael Bourne, Editor, Gee Publishing, 2001
Part 1: The Strategic Planning Process
The objective of a strategic planning process is to identify opportunities where the organization's current or potential capabilities can be successfully and sustainably matched against the needs of its various stakeholder groups. Success is defined by the objective (vision and mission) of each organization. It is measured by the value that it actually delivers to these stakeholders, relative of course, to that provided by their other alternatives.
In a competitively based society, each of the organization's various stakeholders has choices. The owner's of its capital have the option of selling that capital and investing the proceeds in other organizations that they believe will provide them with a greater return on their loaned financial assets. Employees have the freedom to associate with a different organization where they expect to receive a greater return on the time that they invest. Customers usually have the ability to select a different product, service, or supplier for fulfillment of their needs. Suppliers have the choice of providing the inputs required by the organization when it wants them and at the price it's is willing to pay. And communities can commit their limited resources (land, infrastructure, etc.) to those organizations that they believe will prove to be the greatest asset to their constituents.
A successful organization manages its internal processes in order to win against the competition on these stakeholder battlefields. The strategic planning process distributes the organization's always-limited resources among these concurrent challenges. How it deploys them will determine whether it survives as an entity in order to compete again another day.
I offer the following as a generalized model for a Strategic Planning Process:
Figure 1. The Strategic Planning Process
This process forms a closed-loop system. It operates in a continuous cycle, with neither a beginning nor an end. Since most organizations implementing a BSC already have an explicit or tacit strategy in place, I'll start my description with the step identified as number one in this figure. Also, for simplicity, I'll describe the process in terms of the "customer" stakeholder group and leave it as an exercise for the reader to extend it to its organization's other important stakeholders. They usually include owners, employees, union leaders, suppliers, regulators, communities, etc.
Step 1: Choose targeted stakeholder segments
For decades we have been admonished to make explicit in our strategy the customer segments that we intend to serve as well as those we will leave for others to serve. This decision is sometimes referred to as strategic intent. We do this in recognition of the view that we cannot be all things to all potential customers and therefore must focus our limited organizational resources on those chosen market segments. We will secure leadership in them if we can satisfy their members needs better than our competition. Our level of reward will depend on our market share and the maturity and growth rate of that segment's demand.
To make our decision on target market segments, we must understand the opportunity space (potential market segments) and the competitive environment as well as our own organizational competencies. For much of the later part of the last century, we could rely on normative models to predict our chances of success based on the assumption that cumulative experience (as determined by historic relative market share) was its principal driver. Today we understand that flexibility, agility, and rapid learning are more important competitive advantages given the rapidity of technological change and the increasing contribution of often-volatile organizational knowledge.
Making the wrong initial choice may trigger a doom loop from which there's little chance of recovery. However, rapid cycling of this strategic planning process can quickly lead to convergence to a sustainable competitive position. So let's assume that we have initially chosen a set of related market segments where we have a reasonable chance of long-term success. We'll return to this assumption in Step 9 to determine its validity.
Step 2: Identify their requirements
Each customer segment is characterized by its own unique set of requirements. Either objectively or subjectively potential customers test each candidate supplier against these requirements. They choose the one that comes closest to meeting their aggregate needs. They do not weight each criterion equally, and that is what makes them different from one another. One segment might weight price much more highly than reliability. Another may have just the opposite weighting. Once the targeted segments have been selected, it is important to determine their importance weighted supplier selection criteria.
Step 3: Determine performance gaps (external
By asking our targeted customers how we are doing in meeting their various requirements we can identify our performance gaps. Our hope is to close these gaps in order to maintain or improve our relative competitive position. We recognize that if we do nothing, we are likely to loose ground against more aggressive competitors who are pursuing their own improvement objectives. Performance gaps will differ from one targeted segment to another, so we need to apply this step separately for each of the market segments that we are currently serving or considering.
Step 4: Set stakeholder improvement priorities
Improving requirements that are unimportant to a targeted customer segment is often a waste of precious organizational resources that could better be used elsewhere. It's therefore essential that we focus our improvement efforts on major gaps in important customer requirements. The combination of high importance and low performance is the logical basis for ranking opportunities for improvement. Once we have completed this step, we have essentially generated a Pareto diagram of externally identified improvement priorities - tempered by our own strategic objectives.
Step 5: Link stakeholder requirements to internal processes
Many organizations stop at Step 4. Doing so leaves both the responsibility and accountability for improvement unassigned. They may achieve acceptance of the objective but leave undefined each individual's role in making it happen. Naturally, with this uncertainty, they usually conclude that closing critical performance gaps is someone else's job. Like spectators at an athletic event, they sit cheering in the stands, when they should in fact by out on the field as players in this struggle to win. The key to getting their involvement is the linkage of external improvement priorities to internal processes.
One very powerful view of an organization sees it is a collection of interacting processes whose collective output is the vehicle for creating stakeholder value. At the highest level are the macro processes such as product development, customer acquisition, production, procurement, and human resources management. But processes are fractals. As we look at each of their steps through a virtual magnifying glass, we see imbedded within them similar looking processes ... and within them sub-processes ... and within them micro-processes. Every employ has a daily job in which they execute one or more of the steps that are contained within this hierarchy of value creating activities. Their personal link to the overall goals and objectives of the organization flows with the output of these processes as they cumulatively create more value and the consequent increased stakeholder satisfaction.
Step 5 identifies the relationship of each process within the organization to the key stakeholder requirements identified in step 2. It is the transition step from the external to the internal perspective.
Step 6: Establish process improvement priorities (internal perspective)
Knowing which internal processes drive the various targeted stakeholder requirements (from Step 5) and which of those requirements are most in need of strategic improvement (from Step 4), we are now in a position to set internal process improvement priorities. Once completed, we have identified the focal points for changes in the way those involved should do their daily jobs.
The organization can now concentrate its limited resources on the improvement of those leveraged processes with the knowledge that this will produce the greatest strategic return on the investment of those precious resources. And each individual who spends their time executing those key processes will understand why its improvement will be worth their effort. They'll realize that their help is likely to be critical to the organization's strategic success ... that they are an important link in that chain of critical actions.
7: Establish metrics and goals for the process
improvement priorities - the Balanced Scorecard
In my experience, few organizations today make it through Step 6. Identifying with confidence those critical internal processes whose improvement will have the greatest strategic impact is no easy matter. More and more, they are hidden behind a cloud of complexity and confounded by uncertainty and chaos. But for those who do, they now face several nearly daunting challenges:
• Choosing metrics: What exactly should we measure?
• Setting Goals: How will we define success?
• Avoiding over commitment: Do we have the organizational capacity to do all of it?
Defining measures of the output of a process that relate directly to stakeholder requirements is usually straightforward. But these results metrics are not directly actionable. We need to identify the internal process metrics that are the drivers of the desired improvement in these results. Once we have successfully identified them, we need to set time-based goals. In general, they will be stretch goals: difficult but not impossible to achieve.
I can't imagine an organization in today's world where people are sitting around looking for something to do. Everyone already has a pretty full plate of work. They can only squeeze in a limited amount of time to work on process improvement without adversely affecting the performance of those daily jobs. In other words, organizations have a limited improvement capacity. Asking them to do everything will guarantee that the easy ones, not necessarily the most leveraged ones will get done first. We need to filter the priorities established in Step 6 against this limited capacity. In doing so, we create a cut-list. We can do the things above the cut-line, but we don't have the capacity to do the ones below that line ... at least not right now. Acknowledging our limited capacity diffuses the organizational paralysis usually brought on by over commitment.
To focus everyone's attention on the short list of improvement priorities and goals, we create the instrument that has been called the Balanced Scorecard. It captures the results of all of the proceeding steps on a single sheet of paper. It represents a set of metrics and their associated tangible goals that are the best that we can do in advancing our strategic objectives, subject to our available organizational constraints. In a sense the balanced scorecard is merely a rallying flag for all of the effort that has gone into its creation. It is not an end, but an intermediate means for the strategic planning process.
The resulting balanced scorecard is the organization's guide to its improvement priorities. Because it is rooted in the process view of the organization, it can be easily linked from the corporate level down through the process hierarchy to the teams and individuals that are the only ones that make things happen.
Step 8: Improve critical processes
If it were easy to close the gap between current performance and the improvement imperatives established in the previous step, those gaps would have been closed long ago. Certainly focusing the organization's energy around a few specific objectives is a great help. But there is a wide range of approaches that can be used to address these vital few gaps, and they can lead to strikingly different rates of improvement.
The fastest method is to assemble a cadre of process engineers to fundamentally redesign each key process; but this is also the most expensive way to do it. Using the traditional trial-and-error approach not only takes too long, but its actual cost rivals that of the use of an army of process experts. Fortunately, there is a low cost, high-speed approach that was pioneered in the 1930's by Kepner and Tragoe and refined in the 1960's by Japanese TQM practitioners. This improvement model uses teams of process executors who are trained in the basics of the scientific methodology and spend a portion of their time (typically 5-10%) improving their processes. This has proven to be the best way of closing performance gaps when many processes contribute to them.
I am not in any way suggesting that processes that do not make this list should not be improved. A basic cornerstone of TQM is that ALL processes should be continuously improved and that EVERY employee should spend a portion of their time in those activities. What Step 7 does is set priorities for those improvement efforts. Process teams should focus their efforts on improving those outputs that are directly derived from that step.
Individuals involved in multiple processes should concentrate first on those that lie on this critical strategic path. When teams or individuals do not have a clear role in strategic improvement priorities, they should still spend a portion of their time improving the way that they do their daily jobs. But they need to recognize and accept that scarce resources, such as training and internal and external experts, as well as management attention will go first to those who are working on improving the critical processes.
Step 9: Reassess strategy
Organizational defense mechanisms often mandate that our processes be run open loop. We like to plan and do, but have a natural reluctance to check subsequent results against the original plan and take corrective action based on what we learn from that diagnosis. We find this distasteful because the result of the check process all too often is blame rather than learning.
When I first met Ed Deming he was around 80 years old and often noted that 80% of the root causes of defect generation were the process and only 20% the people executing those processes. Each year that went by, that 80% number seemed to grow by 1%. Ed died at age 94 and the last time I saw him he said: "nearly 95% of the problems lie in the process, not the people." I wonder what he would have said had he lived to be 100?
Once we outlaw blame as a management reaction and replace it with constructive learning, we can hope to continuously improve the strategic planning process itself. That is the purpose of Step 9. Before reaching this step, we have identified exactly what we need to do in order to achieve our strategic objectives. We have focused every bit of our available organizational capacity on those required actions. We now ask, "Did we get the results we planned for, and if not, why not?" Out of this diagnosis we can understand weaknesses in our strategic planning process and make improvements for the next cycle. In doing so, we are learning how to plan and act more successfully, and that, after all, is what this is all about.
One sobering result from this reflection step may be that we are doing the best that we can, but we do not have the organizational capacity to do what is necessary in order to achieve our strategic objectives. Often this is the result of competitors who have greater organizational capacity or process know-how, so that although we're improving, we're inevitably loosing ground to them. This painful knowledge should prompt us to seek other competitive niches were we have a chance of winning or face up to the unpleasant reality that our owners remaining equity might best be used by them in some other endeavor. Since very few organizations use their process improvement capacity to their highest strategic advantage, mastery of all of these nine steps has the potential to produce some really unexpected, dark horse winners in the ever-present competitive race.
Dealing with today's strategic planning reality
I'm fascinated by the current notion, often promoted by self-serving consultants, that there's a simple, secret formula for developing a good strategy and it's called a balanced scorecard. "Buy our BSC software," "Attend our BSC seminar," or "Retain our BSC team of experts" and in a few short weeks or months you'll have a winning strategy." And the evidence does seem to suggest that Abe Lincoln may have been right: "... you can fool (nearly) all of the people, some of the time..." But the truth is that developing and implementing a successful strategy still is a very difficult challenge. Their are several contributing factors:
• increasing real-world
• nonexistent data,
• chaos and uncertainty,
• getting organizational commitment and buy-in.
By any measure, organizational life is getting more and more complicated. Everything seems to be both interconnected and important. Clear visions of the future are obscured by this complexity and each group within the organization tries to see through that cloud with their own uniquely colored glasses. The ideal solution - fact based knowledge - is becoming both expensive and time-consuming to generate. In many instances, the important things "are both unknown and unknowable" to quote Ed Deming. We live in a period of unprecedented change. The future is increasingly unpredictable as wave after wave of technological, sociological, and political change break over us. It is a truly exciting time to live in, but an equally frustrating time for strategic planning.
Finally, as organizations transform from physical labor to knowledge based, employees are less willing to simply do as they are told. They need to be enrolled in the strategy before they will work hard to make it happen.
Given these formidable challenges, how can an organization maximize its chances of developing and implementing a winning strategy? Notice that I said "maximize its chances," not guarantee its success. That's the best that any organization can hope for given its tumultuous environment. Here's my advice:
• Take every feasible opportunity to expose employees
first hand to that environment and make sure that they share what they
learn with others within the organization.
• Maximize employee involvement in the strategic planning process itself, by assuring that those with the best knowledge contribute to its relevant steps.
• Use tools that can analyze "fuzzy data", which often is in the form of sentences rather than hard numbers.
• Seek group gut feel, rather than that of individuals who may be distant in both time and intimacy with the current situation.
• Make strategy development an open rather than a secret process within the organization
Sure, there is a risk that by running a wide-open, highly visible strategic planning process a competitor may learn something that they can use against you; but that danger is grossly exaggerated. In reality that risk pales compared to the cost of poor internal alignment caused by a strategy hidden behind a shroud of secrecy. All employees have a "need to know" if they are to contribute effectively to the organization's success.
How an organization executes this 9-step strategic planning process will greatly influence its probability of success. At one extreme, members of the strategic planning department can sit around an isolated table and talk through each of the steps to come up with a scorecard and its associated metrics and goals. In my experience, that approach has a low probability of producing a decisive scorecard and a convincing call to action to those whose efforts are needed to make it happen. At the other end of the practical spectrum, the strategic planning function can orchestrate a broad based effort that synthesizes both internal and external knowledge into a compelling and actionable plan.
In doing so, they will encounter difficulty in processing all of the information and opinions that are generated unless they use some framework and an appropriate toolset for drawing actionable conclusions from the resulting maize of information. That's the purpose of Steps 1a, 2a, and 3a in my model. By numerically weighting the strategic importance of the various stakeholder segments, each segment's hierarchy of requirements, and their perception of our performance on each of their important ones a list of improvement priorities can be generated that separates Juran's "vital few" from his "important many." Part 2 will expand more on these "a" steps.
In this, Part 1 of the article, I have described a 9-step framework that I believe represents a comprehensive process that has as one of its many important outputs a set of balanced scorecards that deploy strategic goals down to the action agents that really make strategy happen. In the next two parts, I will describe in detail the actual methodology that I use in implementing Steps 1-6 (Part 2: Setting Process Improvement Priorities) and Step 7 (Part 3: Selecting Scorecard Metrics). Step 8 is the theme of my Process Management Model.
Part 2: Setting Improvement
I'm a long-time advocate of the KISS principle: "Keep it simple, stupid," or its more formal ancestor known as Ockham's razor. But as problems become more complex, so unfortunately do their simplest solutions. Scan ahead in this part and your initial reaction may be that what I'm proposing looks awfully complicated. But, if there's a simpler way of getting to a truly effective answer, I've yet to find it; nor am I aware of anyone else who has.
That's because one of the inevitable consequences of our current form of progress is that over time it creates ever-increasing complexity. We can no longer manage that complexity with the basic toolset that worked in a simpler, bygone era. Those tools helped in understanding systems where the whole effectively behaved as the sum of its individual parts. The tools were used to break a big problem into a set of small, manageable pieces. By optimizing the pieces, we could expect to optimize the whole system. The very best of mangers could even do this in their heads.
Today, complexity arises from the increasing interdependencies between the many small pieces of a big issue. The response "it depends" that once served as a ubiquitous excuse, now takes on legitimate meaning. The interdependencies become further compounded by their eventual non-linearity. Together these two effects have pushed the critical problem space well beyond the capabilities of simple tools and individual gut feel. More and more often we are confronted with situations where the whole is much greater than the sum of its individual parts. The setting of process improvement priorities now resides in that elusive domain. Yet it is essential to identify the real improvement priorities, not just for the effective use of limited organizational change capacity, but also to weave the convincing story needed to marshal organizational support and buy-in.
Even when an insightful executive can see through that cloud of complexity, verbal explanations are ineffective in transferring his gut feel to others. They must take his conclusions on faith. But today, fewer and fewer organizations can rely on faith as their alignment mechanism. Knowledge workers in particular demand a compellingly and logical argument before they will sincerely commit to "making it happen."
In 1979 the Japanese Union of Scientists and Engineers, the driving force behind Japan's TQM revolution, codified a set of tools that they called the 7-Management and Planning Tools (or 7-MP). Over the last thirty years the 7-MP have proven their effectiveness in the achievement of consensus or what we might call "collective" or "group gut feel." It's one of those tools, the Matrix Diagram, which I will be using here.
Other tools useful in dealing with this increased complexity have been around for half-a-century. The challenge is to choose the simplest of these tools that can adequately address the issue at hand. Oversimplifying the problem in order force-fit it to our more familiar approaches can only create the illusion of understanding, which cannot be a sound foundation for action. So be forewarned that what follows, in my view is the least complicated way of correctly identifying strategic process improvement priorities in today's increasingly complex environment.
This Part describes a methodology for deriving process improvement priorities from an organization's strategy. It relies heavily on the framework used in Quality Function Deployment 1 (QFD). That framework uses a series of interrelated matrices to numerically define the strength of the causal relationships that exist between the "what's" and "how's" of effective planning. As you will see, it significantly extends the use of simple casual-loop diagrams (as used for example in BSC Strategy Maps) that only serve to identify major causal linkages. By quantifying the strengths of these linkages and providing an aggregation mechanism, this approach often uncovers pervasive process improvement opportunities that would be missed when only the most obvious dependencies are considered. Furthermore, since its output is a numerically weighted list of strategic process improvement priorities, it helps us get the greatest strategic bang for the organization's limited change capacity buck.
We will start by looking at various strategies and their relationships to segmented stakeholder requirements. This will allow us to place a strategically chosen "importance" weighting on each requirement. In doing so, we explicitly identify the specific stakeholder segments that we choose to serve and by implication, those that are not on our strategic agenda. Next, we will determine actual performance, both absolute (based on customer needs and wants) and relative (based on competitor performance) and combine strategic importance and performance to generate a numerical scoring where the higher the value the greater is the strategic need for improvement of that particular stakeholder requirement.
Our second matrix defines the relationship between stakeholder requirements and each of the organization's various value creating processes. It quantifies the impact of each key internal process on each of the stakeholder requirements. Finally, we will combine improvement priorities derived from the first matrix with process linkages from the second to produce a process improvement prioritization list. This list will represent a scored ordering of processes in need of improvement in terms of the impact of these improvements on stakeholder satisfaction and, therefore, strategic success.
As I will show, this approach is amenable to various levels of detail. At one extreme, it reduces to a simple normative model that states "if this is your strategy, than this is what your targeted stakeholders expect and these are the processes you have to get right in order to satisfy those expectations." For simplicity, that's the example I'll use here. At the other extreme, detailed studies may be necessary to determine the organizations real vs. professed strategy, actual customer requirements by targeted segment, perceived performance, organizational barriers, etc. Where in this spectrum a particular situation lies depends on the level of detail necessary to achieve the required consensus for action. Often this is determined through a process of successive approximations, starting with the simple normative model and adding more detail until that consensus is reached.
One definition of consensus is the achievement of a state in which the least supportive member of the group "can live with" the majority's view. But a consensus for action often requires a much stronger commitment from that last individual, particularly when their active support and participation is required to make that action happen.
Stakeholders and Their Requirements
Organizations have a number of stakeholders. Generally, we identify them as:
• stockholders or owners,
• suppliers, and
• the communities in which we do business.
In some cultures, the environment and future generations are being added to this list (see The Fifth Fitness). In some industries, there are multiple customers. For example, in higher education customers can include parents, future employers, academic peers, and research sponsors, as well as students and alumni. In healthcare not only patients but also doctors, hospitals, regulatory agencies, and insurers needs must be addressed. Where appropriate, distinctions need to be made between historical, current, and future requirements, as well as different "classes" of stakeholders such as large corporations, small businesses and individuals.
An organization must identify its strategy and the key requirements for each of its strategically chosen stakeholders. For example, is its stockholder strategy income, growth or non-profit driven? If it is income driven, then its targeted stockholders will place a high weighting on a steady dividend stream and a stable stock price. They will be satisfied with average returns on their investment. On the other hand, the stockholders of growth driven companies do not value dividends, accept above average price volatility, but demand strong long-term growth in stock price. They expect to be compensated for higher volatility (or b) with above average long-term returns. The owners of non-profit organizations usually have non-financial expectations for the return on their investment.
Employee related strategies range from nurturing to competitive. Employees in nurturing organizations hope for security, lifetime employment, liberal benefits, low stress and a family-like environment, while those in internally competitive companies seek an entrepreneurial environment with rapid personal advancement opportunities. They place much higher value on short-term rewards than on long-term job security.
Obviously, the various stakeholder strategies need to form a self-consistent set. They are not in general independent. Income driven companies tend to have nurturing employee strategies, while growth driven companies often have more competitive employee strategies.
Strategies and the Treacy and Wiersema Value Disciplines
As you can see from the above examples, the strategy is really a name for a particular profile of targeted stakeholder requirements. The name only takes on general meaning if most companies or business units can be assigned to one of the identified categories based on similarity of their targeted stakeholder requirements.
One such recent classification system is that of Treacy and Wiersema 2 (T/W). They have defined three "Value Disciplines" as a way for classifying companies' customer strategies. In the remainder of this Part, I will be using the T/W model as an example of the application of this methodology. Using their one-dimensional view of the organization's stakeholders greatly simplifies my description of the elements of the methodology. But:
Please keep in mind that the T/W model applies only to customer strategies. All stakeholder strategies must be considered if a robust prioritization is to be achieved. Omission of a stakeholder group often will lead to priorities selected at their expense. For example, the T/W approach alone will probably give the wrong answer if applied to a company whose most important strategic imperative is increased stockholder value through growth. Customers do not usually value the growth of their suppliers. Therefore, revenue growth generating processes will tend to be de-emphasized when only the customer perspective is taken into account. So in applying what follows to a particular company situation the T/W Value Disciplines MUST BE augmented or replaced with a similar type classification for the all of the important stakeholder strategies. The methodology for doing this is quite straightforward.
T/W identify three Value Disciplines, which they called "operational excellence," "product leadership," and "customer intimacy":
Companies pursuing an operational excellence strategy provide the lowest total purchase cost to their customers by providing high quality (conformance to specification), low price, and ease of purchase. They accomplish this by streamlining processes to minimize costs and hassle, standardizing, providing high-speed transactions, and creating a culture that abhors waste and rewards efficiency.
Product leadership companies provide the best possible product to their customers. They focus on creativity and rapid commercialization. They relentlessly pursue ways to leapfrog their own products before someone else does. Intermediate milestones, keeping on track, and celebrating interim victories, characterize their product development process. They operate a loose, entrepreneurial organization, are results driven, and encourage individual efforts.
Customer intimate companies provide their key customers with the best total solution to their problem. Their focus is on individual key customers rather than markets. Their most important process is solution development, which is characterized by delegated decision-making and specific rather than general solutions.
Key Customer Requirements
Let's now look from the perspective of customers. They have a portfolio of requirements and will most often choose the supplier that best meets them. There are many ways to define the general set of customer requirements. Often they need to be industry specific. For manufacturing, the set of requirements I usually use is as follows:
a. Performance Specifications. These are defined by the performance characteristics of the product relative to competition. Often they relate to speed, accuracy, resource usage, size, etc.
b. Fitness for use. Does the product do what I need to have done?
c. Fitness for latent needs. Does the product meet an important need that I did not previously know I had?
d. Aesthetics. Is the product visually appealing?
a. Conformance to specification. Does the product actually perform as specified when received?
b. Reliability. Does the product continue to perform as specified over its useful life?
c. Durability. Is the product robust to normal wear and tear?
d. Serviceability. Is the product easily serviced when needed?
a. Price. This is the actual realized selling price, after discounts, etc.
b. Cost of ownership. The additional life-cycle costs I incur with the product including inspection, inventory carrying costs to cover poor delivery, rework costs, warrantee costs, etc.
a. Quoted Lead Time. Ability to get a commitment to receive the product when I want it.
b. Minimum/maximum order size. Ability to get the product in the quantity that I need.
a. Delivery. Past performance to committed delivery dates.
b. Responsiveness. Broadly defined, this is the ability to get timely answers to all queries.
a. Willingness to partner.
In any particular situation it is important to replace the above list with an appropriate classification of key customer requirements. These requirements answer the question: What do our customers consider in making their purchase decision between alternative products and/or suppliers?
Relating Strategy to Key Customer Requirements
If we consider customers using the above purchase criteria, and map them against the T/W Value Disciplines, we arrive at Figure 2.
Figure 2. Relating Strategy to Customer Requirements
The central part of this matrix arrays the three Value Disciplines against the list of possible customer requirement. The symbol used at the intersections represents their degree of relationship. For example, the double circle shows that there is a strong relationship between Product Leadership and Specifications. The single circle shows that there is a moderate relationship between Customer Intimacy and Ownership Costs. The triangle denotes a weak relationship between Operational Excellence and aesthetics, etc. Blank cells denote no significant relationship.
Implicit in the use of this tool is that these relationships remain essentially constant over the appropriate planning period, which is typically a year. By regularly revisiting them, the matrix can be updated to better reflect the current situation. Also, for simplicity I have omitted an additional step often used in QFD. In that step, we examine the interrelationships between the various requirements to identify conflicts and reinforcements. We capture them in what are called "roofs" and use them to identify the impact of candidate changes in one selected requirement on the others. This becomes necessary when the improvement of one requirement can worsen performance on another. For example, adding features may be offset by an undesirable increase in price. There are also synergistic improvements. Quality improvement usually leads to a reduced cost and increased responsiveness. If changing degrees of relationship and linkages between requirements becomes important, I generally abandon this entire approach in favor of System Dynamics simulation modeling since it is optimized for those dynamic situations.
The filled in matrix in Figure 2 represents my interpretation of the operational definitions of the different T/W strategies. For example, the matrix defines a "customer intimate" company as one that sets its highest priority on providing products and services that meet customers needs, including latent needs, while being both responsive and willing to form collaborative relationships. Furthermore, it makes sure that it has competitive specifications, low post-delivery quality and ownership costs, and that its reputation is consistent with these goals. Finally, it ensures that delivery and minimum order size do not conflict with its higher priorities. Its customers are indifferent to the blank requirements unless performance drops below an easily maintained level.
Once filled out, the matrix becomes the dictionary that defines the various strategies. As you look across each row, you can clearly see that each strategy has its own distinctive signature. Should a new customer segment appear that has a significantly different set of key requirements, a new name must be created and added to the list of strategies to capture that unique segment.
In filling out the matrix, I have adhered to some simple pragmatic rules. For it to be useful, the matrix should be sparsely populated. There is a tendency for people to see strong relationships between all of the elements. If this happens, than the matrix looses its ability to distinguish the different strategies. When working with a group of people, a facilitator can help by asking questions such as "what is the most important relationship?" or "where is the relationship very weak or insignificant?" or "which is more important ‘a' or ‘b'"? A good goal is to have 40% - 60% of the elements blank and a fairly uniform distribution of strong, medium, and weak symbols. Looking along both rows and columns, there should be significant differences in the degree of relationship. In other words, the strategies should look different from one another. The use of a non-linear weighting scale will further help in combating too many unimportant relationships.
In developing or refining a matrix, a team may encounter significant disagreement about a relationship. If progress is to be made, the team should make a tentative choice. It can then go back after completing the exercise to test sensitivity of the conclusions to that particular relationship. This is made easy through the use of QFD specific or spreadsheet software. I recommend QualiSoft's QFD Designer, which I used to prepare Figures 2 and 3. Usually many relationships have to change significantly for it to make any difference in the overall conclusions. If sensitive relationships are found, than further study of them is required. For example, if improvement priorities change depending on how important reliability is to customers, than a small focused survey can be done to answer that specific question. Consensus and buy-in are essential parts of this process and can only be achieved by bringing actual data to significant areas of disagreement.
There are two alternatives for the next
step. If the organization knows which of the three strategies it is
following, then "1" is used in that strategy's column entry and "0" is
entered for all of the others. The "importance to customer" row is
calculated by replacing the symbol in each matrix element with the
numerical weight for that symbol, multiplying by the number in that row
of the "strategy" column, and adding the resulting numbers by column.
In this case, the result would simply be the weights for the chosen
However, the organization often determines that its business is or should be split among the three value disciplines, say 70%-20%-10%, and that its internal processes do not differentiate between orders from customers in different segments 3. In this case the "strategy" column would contain the numbers .7, .2, and .1 (always totaling 1.0) and the same calculation would be made to determine overall importance to customers. Our purpose here is to discount important requirements for the less strategically significant customer segments.
The particular weights chosen here, 9-3-1, are used to accentuate the differences in relationships. This is a common set used in QFD. Others include 5-3-1 and 3-2-1. Again, sensitivity testing using different weighing schemes can determine the robustness of the conclusions. What is really important is that items toward the top of the list really belong there and visa versa.
Sometimes, the organization cannot agree on which value discipline(s) it is following. This may result from lack of data, multiple strategies, or inappropriateness of the strategy classification system to their particular business. In this case, a second approach may be necessary: a market segmentation study. One way of doing this is through surveys or interviews of a representative sample of key customers (50-100). This sample can include past, present and potential future customers and non-customers (i.e., customers of competitors). Each customer is asked to distribute 100 points between the key customer requirements.
It is also useful to uncover trends in their point allocations by asking for significant differences in how they would have distributed the points five years ago and what they think might be requirements of increasing and decreasing importance over the next five years (remember, the total stays at 100). For example, point allocations to quality and delivery have tended to drop, as they have become "givens" for doing businesses, while relationship, JIT delivery, and e-commerce are likely to increase in importance in the future. At the same time, need for improvement of the organization and its principal competitors can be ascertained using a scale of zero (low need) to ten (high need) for later use.
The resulting data are sorted into groups of customers having similar key customer requirements. This can be done using statistical sorting techniques or by subjective means. I prefer the latter. Translating the point allocations into bar charts and laying them out on a table, they can be visually grouped into similar profiles or customer fingerprints. Occasionally, an organization might require more rigorous analysis although in my experience the increased expense adds little or no real value.
It is worthwhile to mention here the techniques developed by Noriaki Kano4 for distinguishing requirements that are "delighters", "satisfiers", and "must-be's (without it, they are dissatisfied)." This simplified form of conjoint analysis is widely used in Japan.
Often, industry surveys published in trade journals or analyst's research reports can be used in place of, or as a adjunct to direct surveys or interviews. This reduces the cost of determining key customer requirements but at the price of customer specificity and interactive learning through the interview process. Either way, the result is a direct numerical scoring of key customer requirements by importance to them (the higher the points, the greater the importance).
The resulting numbers for a specific customer segment are entered into the "importance to customer" row of the matrix. This time the calculation is run in reverse, multiplying the weights by the importance, and now summing the result across the rows and entering the sum into the "strategy" column. Ideally, one of the numbers in the strategy column will be much larger than the others. This represents the appropriate Value Discipline being followed. If there is no clear "winner", then the T/W model is not useful for this market segment. What we have in fact done is used the methodology as a diagnostic to determine the appropriate strategy name based on key customer requirements. If the T/W names don't fit, then we can give the new profile its own, unique name.
Assessing Need for Improvement
The objective to this point has been to rate the key customer requirements in terms of importance to customers in the targeted market segment. We did this by using the appropriate T/W Value Discipline or by direct measurement. The next step is to determine need for improvement. I'll be assuming that the product of "importance to customer" and "need for improvement" is a good indicator of "improvement priority." For those who are unsettled by this assumption, I refer you to the emerging branch of mathematics known as "fuzzy logic." A more rigorous approach would be to use the utility function from economics theory, but that would represent a much more complicated refinement.
Here we have three alternatives:
1. By entering "1" in the "need for improvement" row, we are in effect determining the key customer requirements you need to get right in order to satisfy those customers. In the next step, this will produce the enabling business processes or core competencies required to achieve leadership in this strategy or Value Disciple.
2. By entering absolute need for improvement in the "need for improvement" row, we are in effect determining the performance gap relative to customers perceived needs. This will lead to a prioritization of improvements most useful to the market leader in maintaining or increasing its leadership position. There are two sources for these data:
Consensus voting by knowledgeable insiders.
b. Direct data from customers. For example, if we asked customers to rate our performance on a scale of one to ten, where ten would be their ideal supplier, then the difference between our score and ten would be an indication of our absolute need for improvement on that requirement.
3. By entering relative need for improvement in the "need for improvement" row, we are in effect determining the performance gap relative to our best competitor with respect to that requirement. This will lead to a prioritization of improvements with the objective of gaining share against the market leader. Again, there are two sources for relative performance data:
a. Consensus voting by
b. Direct data from customers. For example, customers can be asked to rate our performance relative to each competitor on a scale of one to ten. The numerical difference between us and the market leader, or the best in class for each requirement can then be used as a measure of "need for improvement."
"Need for improvement" scores can be determined in this way depending on the prioritization objective, be it:
• "what do we have to get
• "what do we have to do to maintain leadership?", or
• "what do we have to improve in order to gain market share?"
This is also the place where trend data can be used to explain past performance and to predict future areas in need of improvement.
It is "nice" to have the importance to customer row total 100 and the "need for improvement scores" be based on the original range of from zero to ten. This can be accomplished be re-normalizing and rounding-off the entries where necessary.
Linking Customer Requirements to Business Processes
We can now turn to our second matrix. This matrix relates the key customer requirements to the underlying business processes. There are many ways to classify business processes. The one I will use here is the system described by Tom Davenport 4. We will use the requirements improvement priority weights determined in the previous matrix. Our objective is to identify the impact of each business process on each of these key customer requirements. Following the same rules as previously described, figure 3 represents my view of these relationships.
Figure 3. Linking Requirements to Processes
This matrix contains the essence of an organization's understanding of its business processes. It is probably unique to a given industry and market segment. In its detail, it may be dependent on each individual organization. In a sense, it captures the organization's knowledge of the internal drivers for customer (or stakeholder) satisfaction. When done by a group of process experts, it constitutes their collective wisdom as to the key business drivers in their particular industry. It is the truly proprietary part of what an organization learns about itself in applying this approach.
One of the most important properties of this matrix is that it is not diagonal; there is not a unique one-to-one correspondence between a key customer requirement and a single business process. Consider, for example, on-time delivery. Businesses do not usually have an on-time delivery process, staffed by an on-time delivery department and led by a Vice President of on-time delivery. On-time delivery performance depends instead on many independently managed processes within an organization (see for example my article on "Metrics for the Order Fulfillment Process"). In figure 3, the major drivers are manufacturing, logistics (supplier delivery), and information management (scheduling and MRP). It is this multiple-dependency that creates an interconnected business "system," which in turn causes the need for this approach to prioritization.
Once the matrix is complete and the customer based improvement priorities transferred from the first matrix, the initial priority can be calculated. This is done by multiplying the weights by the improvement priority and summing the columns. But before the final improvement priority is determined, the issue of degree of difficulty or organizational readiness must be addressed.
Processes differ in complexity, both from a technical and people perspective. Improvement is more difficult in a process where the root causes relate to human behavior then it is for a process where only equipment or methods need to be changed. Also, data provides the basic fuel for the improvement process. Can the needed data be generated by the improvement team or does it have to come from someone else? Cross-functional processes can be complicated by conflicting objectives and ever-present politics. Since our goal is rapid improvement in results, we need to raise the priority of processes that can be improved quickly and drop the priority of the more difficult ones. We do this by adding the row titled "organizational difficulty" to the matrix.
One very interesting commonly observed phenomenon is that "success breeds success." Over time, many of the initial organizational barriers dissolve on their own, making the passed-over process improvements more easily tractable. Often, the elimination of the old culture of blame is the key to this transformation.
Organizational difficulty is characterized using a subjective scale ranging from "1" (low) to "5" (high). In practice, teams can easily assign values, since the consideration becomes the number and severity of issues rather than who is at fault. Once the organizational difficulty is established, the final priority for process improvement is determined by dividing the initial priority by the organizational difficulty and rescaling.
The QFD Designer software includes a bar graphing capability that makes the final results for each matrix quickly apparent. The use of the symbols rather than numbers in filling out the matrices serves a similar role in the visual display of the relevant information.
The final step in completing the matrix is to determine principal performance metrics and their associated goals, at least for the high priority improvement targets. These goals must be aggressive yet achievable. When met, they would move this process from its current high to a significantly lower priority for improvement. It is these performance metrics and goals that have earned their place on the appropriate BSC.
In addition to my writings on the half-life method for goal setting, Part 3 will describe a systematic approach for identifying the appropriate measures and metrics for each of the resulting strategic processes improvements.
Results for the Normative Model
Figure 3 has been completed for an organization
successfully pursuing operational excellence. The improvement
priorities were determined based on customer requirements rather than
Organizational difficulty was assumed the same for all processes. Principal performance goals are based on an organization that is delighting its customers (i.e. there's no customer identified need for improvement). The resulting process priorities are score ordered in figure 4 in terms of decreasing priority.
Business Process Raw Score Normalized Score Cumulative %
Manufacturing 466 21 21
Identification 249 11 33
Integrated Logistics 234 11 43
Management 205 9 53
Customer Acquisition 178 8 61
Product Development 177 8 69
Asset Management 175 8 77
Performance Monitoring 155 7 84
Information Management 141 6 91
Post-Sales Service 75 3 94
Order Management 74 3 97
Planning and Resource
Allocation 57 3 100
Figure 4. Process Priorities for Operational Excellence
The normalized scores are calculated by dividing the raw score by the total of all raw scores and than multiplying by 100. In can be interpreted as the percentage of effort or resources that should be focused on maintaining that process at superior performance levels. It should serve as a major input into an organization's budgeting and resource allocation processes. The last column represents the cumulative normalized scores.
As can be seen from figure 4, the number one priories of an operationally excellent company are its manufacturing related processes. Understanding its customer requirements and managing its suppliers are next in importance. Getting these three processes right will get them nearly half way there.
Following the same procedure as above, figures 5 and 6 show the process priorities for product leadership and customer intimacy.
Key Business Process Raw Score Normalized
Score Cumulative %
Identification 216 25 25
Product Development 136 16 41
Post-Sales Service 123 14 55
Human Resources Management 105 12 68
Customer Acquisition 81 9 77
Planning and Resource
Allocation 72 8 85
Asset Management 39 5 90
Manufacturing 37 4 94
Information Management 30 3 98
Integrated Logistics 12 1 99
Order Management 7 1 100
Performance Monitoring 1 0 100
Process Priorities for Product Leadership
Success in product leadership depends heavily on understanding customer requirements. In fact it's more important than the product development process itself. This result is entirely consistent with the TQM admonition: "market in, not product out." Next in importance are product development, post-sales service, and HR management. Post-sales service is important because I assumed that it played a major role in determining fitness for use, a very important customer requirement for product leadership. HR management is key in attracting and retaining the creative people needed for product leadership.
Key Business Process Raw Score
Normalized Score Cumulative %
Customer Requirements Identification 349 19 19
Customer Acquisition 304 17 36
Post-Sales Service 261 14 50
Product Development 189 10 61
Planning and Resource Allocation 156 9 69
Human Resources Management 147 8 77
Manufacturing 135 7 85
Information Monitoring 111 6 91
Performance Monitoring 54 3 94
Order Management 48 3 97
Asset Management 35 2 99
Integrated Logistics 25 1 100
Figure 6. Process Priorities for Customer Intimacy
Winning in customer intimacy requires excellence in all processes that directly touch the customer. Most important are understanding their requirements, acquiring and retaining them, and maintaining high levels of post-sales support.
Conclusions for Part 2
At the start of this Part, I said that this approach is amenable to various levels of detail. The examples used here are at the simplest level and provide a normative model for process prioritization based on Treacy and Wiersema's Value Disciplines (figures 4-6). There are no real surprises in the normative model, and that's good news. The methodology passes this simple validation test.
The rich and counter-intuitive insights arise when actual strategies, stakeholder requirements, performance, and constraints are added to the picture. But unlike individual gut feel, how these collective conclusions were reached can be explained to others by following the logic trail. After stripping away what turns out to be the unessential elements of the two matrices, a much simpler picture unfolds, one that is easily used to illuminate that logic path. I refer you to Analog Devices later version of its Scorecard Story for such an example.
*This Part is an extension of a research project done for a major international consulting company in 1995 and described in a working paper that I wrote that year.
1 See for example: Yoji Akao (Editor), "Quality Function Deployment: Integrating Customer Requirements into Product Design", Productivity Press Inc., May 1990, ISBN: 0915299410
2 Michael Treacy and Fred Wiersema, "The Discipline of Market Leaders: Choose Your Customers, Narrow Your Focus, Dominate Your Market", Addison Wesley Longman, Inc., 1994, ISBN: 0201406489
3 In this case, the possibility of creating "cells" within a process that are dedicated to a particular customer segment should be investigated.
4 See for example: Shoji Shiba, Alan Graham, and David Walden, "A New American TQM: Four Practical Revolutions in Management", Productivity Press Inc., January 1993, ISBN: 1563270323, pg. 221.
5 Thomas H. Davenport, "Process Innovation: Reengineering Work Through Information Technology", Harvard Business School Press (October 1992) ISBN: 0875843662
Part 3: Selecting Scorecard Metrics*
A balanced scorecard contains a concise set of strategically important measures. They capture the vital few drivers of the organization's future success. I've called these scorecard measures "metrics" and defined them as:
"Metrics are a subset of measures of those processes whose improvement is critical to the success of the organization"
Once we have identified those processes, we face the challenge of selecting this subset from a seemingly endless list of possibilities. Usually this decision is based on what measures are already available or can easily be obtained, benchmarking studies, or executive edict. But there is a much better way of doing it.
Measures of a process come in two flavors: I call them "results measures" and "process measures," although each has many aliases:
Results Measures Process Measures
Whichever set of names you choose, there is a very important difference between them:
Results measures characterize the output of the process. They are the consequences of actions taken within it. Since they are descriptors of the output, they relate directly or indirectly to things that a customer of that process can sense or measure.
Process measures, on the other hand, are the internal measures from within the process that determine these results. In most cases, the customer has little or no interest in or knowledge of them.
The SIPOC Method
One very useful model for generating candidate measures is called the SIPOC method. SIPOC stands for
In using this model, we usually start by identifying all of the customers of the process and determine their complete sets of requirements. Here, customers include both the external purchaser of the final product or service as well as other internal processes that are part of the organization's value creating activities (or, as we say in TQM: "The next process step is the customer."). Through a process called "Voice of the Customer" we translate these requirements into results measures that characterize the output of the process in terms that are both meaningful to and measurable by the process executors. This translation is necessary because the customer often describes their requirements in words that do not have a direct process counterpart.
Next, we reverse this
procedure by identifying all of the external inputs that we need in
order to execute the process, define our requirements for each of these
inputs, and ideally working with our suppliers, translate them back into
a set of specifications that are expressed in the supplier's own
language ("Voice of the Supplier").
Output measures and their associated quantifiable customer requirements (Output®Customer) are clearly results measures. Measures associated with steps internal to the process (Process) are obviously process measures. But what about input and supplier measures (Supplier®Input)? Symmetry would suggest that since they are results measures of the supplier's value creation process, they must also be results measures for our process. But is that necessarily so? In other words, can a measure that is a results measure for an upstream process be a process measure in a subsequent step? The answer here is a little bit tricky.
What is different about Supplier®Input measures are that we cannot improve them directly from within our own process. We can only do so indirectly by changing specifications, or suppliers, or through the redesign of our product and/or process ("design for x-ability"). Their actual improvement is directly controllable only by the supplier of that input. Often we have a limited ability to affect our supplier's control or improvement efforts (through partnering, for example) or to redesign our products and/or processes. If that is the case, then we need to treat that measure as a given (that is, a constant) and that measure's classification into the results or process category then becomes moot.
Generally speaking, to indirectly change an input measure requires the exercise of different internal process within our organization - the supplier selection process by which we choose suppliers, and/or the product and process design processes. Even in that case, it is difficult to argue that they are anything but results measures. In other words, unless we include within our process sub-processes for supplier selection and product/process redesign, we must view these measures as the result measures of other internal or external processes.
Any given process is part of a system of interacting processes. This is one of the important reasons why it's critical to have sponsorship of all improvement efforts by someone who is in a position to set appropriate boundaries and constraints to that effort.
The Math of Metrics
a mathematical point of view, the last alias-pair is the traditional
choice of terms. For each results measure, we can write a symbolic
equation that relates this dependent measure (or more correctly,
"dependent variable") to the independent ones:
In words, this equation simply states that the dependent measure, yi, is a function of (i.e. depends on, or is determined by) all of the independent measures: x1, x2, up to xn, where n is the total number of independent process measures.
For example, if the process were baking a cake, then one dependent measure would be the "lightness" (in the Language of the Customer) of the resulting cake, measured by its density (in the Language of the Process) in grams per cubic centimeter. Here, y1 would be cake density and the goal is for it to be in a specific range: not too light, not too heavy, but just right. What about the x's? The list would include oven temperature, cooking time, amounts used of the various ingredients, freshness of ingredients, etc. These are the measures that are included or implied in a clear recipe (or Standard Operating Procedure (SOP)). Other dependent measures would include moistness, sweetness, and flavor for example and we could create instruments that would measure each of them, as well as establishing each of their associated target ranges. Each dependent measure would depend on one or more of the many independent measures.
Determining drivers of change
In general, we are trying to limit variation
of and/or improve dependent measures in order to make our product more
attractive to its customers. So let's look at how this equation changes
with changes in the independent measures:
The symbol "" stands for a small change in the measure. So this equation says that the change in a dependent measure is the sum of the weighted changes in all of the independent measures. For very small changes in the measures, mathematicians can show that this simple additive relationship holds in most practical cases. The weights aij are sometimes called "influence coefficients" or "impact parameters." 1 They represent the effect that a small change in the jth independent measure has on the ith dependent measure. If aij is zero, than small changes in its independent measure have no effect on that dependent measure. If the value of aij is large compared to the other coefficients, then the dependent measure is very sensitive to changes in that independent measure. It's these influential independent measures that are usually the targets for both process control and process improvement and are therefore candidate scorecard metrics.
In process control, they are called "critical nodes." By "locking" them, we assure that variation in the dependent measures that they affect will be maintained within a range that's acceptable (but not necessarily satisfactory) to the customer. For process improvement they indicate the likely root causes of the gap between current and target results.
Unfortunately in practice, for large changes in the measures, this simple model is often limited by two phenomena: "non-linearity" and "interaction." Non-linearity causes the influence coefficients to change (increase or decrease) for large changes in their independent measure. Interaction occurs when interdependencies develop between the various independent measures (they loose their independence).
Some Simple Examples
The exact mathematical function takes on different forms for different dependant measures and processes. Here are some examples:
The time required to execute a process from its start to its
finish is called its cycle time. If the various x's are the cycle
times, tj, for the internal process steps that lie on the "critical
path," then the total cycle time, T, is
overall yield of a process depends on the sequential yield of the
internal sub-process steps. Let's say that if the process were perfect
(no internal yield loss), it would produce 100 output units. If the
actual yield in the first step is 90%, then only 90 potential outputs
survive it to the next step. If that step's yield were 80%, then only
80% of those 90 or 72 would make it to the next step, etc. Therefore,
the overall yield is given by:
For Example 1 above, all of the influence coefficients have a constant value of 1, that is any increase or decrease in a critical path cycle time simply adds or subtracts that change from the total cycle time. We could include non-critical path sub-process cycle times, but their influence coefficients would all be zero (until they became long enough to enter the critical path). On the other hand, for example 2 it is straight forward to show that the influence coefficient is inversely proportional to that sub-process' yield (aTj=YT/Yj). What this means is that improving a low yielding process step by 1% (for example from 25% to 26%) has a greater impact on total yield than that same 1% improvement in a high yielding process step (going from say 95% to 96%). In other words, lower yield process steps have larger influence coefficients.
In many manufacturing environments, process or manufacturing engineers know the mathematical relationships between the dependent and independent measures. Usually they do this based on a physical or chemical theory of what's happening in the process. When this is the case, these experts can help in the selection of those independent measures that are the principle drivers of change in any given results metric. Once identified, these process metrics generally represent the primary targets for improvement efforts and are tracked on the appropriate scorecard.
Empirically Determining Process Metrics
As a rule-of-thumb, low influence coefficient independent measures vastly outnumber the critical few (see Why Do Root Cause Analysis?). So trial-and-error is not a viable option. Finding the process metrics in practice often ends up requiring a mixture of both art and science.
When a theoretical equation does not exist or is not known, we need to resort to empirical observation. Total Quality Management (TQM) employs teams that apply the scientific methodology (the PDCA Cycle and the 7-Step Method) and basic analysis tools (the 7 QC Tools) for identification of the root causes (process metrics) of undesirable outcomes (results metrics). I've explained this process in more detail in my article "Are There Limits to TQM?"
The vast majority of process improvements can be discovered using these simple scientific tools. For more complex situations, three additional approaches are sometimes used: heuristic techniques, design of experiments (DOE), and simulation modeling.
I once assisted a team trying to reduce defects in welded pipe used in the oil industry. The particular defect was called "hook cracks" since they had the shape of a fishhook. In stratifying defect data by shift, I discovered that one crew had significantly lower defect levels than the others. I narrowed it down to the welder operator and interviewed him in the hopes of documenting his "secret" so that this best practice could be shared with the others. Each welder setting was specified with a range determined by the industrial engineers. I asked him how he chose a setting from within these ranges and his answer was "I can tell by the sound the welder makes." The other operators just tried to pick the mid-point. The IE's response: "Sound has nothing to do with weld quality."
A few months later I visited an identical pipe mill in Japan where the operators relied on an additional meter to adjust the mill settings. Using a microphone placed near the weld site and connected to a measuring instrument (a spectrum analyzer), their IE's had determined that if the sound frequency was within a certain range, a perfect weld was produced. Outside that range, the resulting product was defective. What was the defect? No one remembered at first since the discovery had been made several years before. Finding an old-timer they came back with the answer: "Something called hook cracks." Why should a good weld have a certain pitch to the sound it made? There was no accepted theoretical explanation; it simply worked. The Japanese IE's were willing to accept this heuristic observation while their American counterparts had discarded it as scientifically baseless.
As another example, Kano 2 observed an important non-linearity in the independent measures that we call customer satisfaction. He classified the independent attributes that drive customer satisfaction (such as particular product features, price, availability, reliability, etc.) into four categories:
Independent Measure type Influence coefficient Implication
One-dimensional (satisfier) Constant The more the better, the less the worse
(delighter) Zero below a threshold, Absence does not dissatisfy,
then positive and increasing presence produces significantly greater
with increasing x satisfaction
(dissatisfier) Negative and decreasing with Presence does not
increasing x to a value absence produces significantly greater
of zero at a threshold dissatisfaction
Neutral (indifferent) Relatively small or zero Don't care whether it's there or not
To place each independent measure into one of these categories, Kano developed a structured multiple-choice survey tool. He than created a heuristic "decoder ring" for determining the measure type from the responses to paired questions. By understanding current performance and the type of measure, the user could than rank all of the independent measures by their improvement's impact on customer satisfaction.
In general, heuristic methods are based on empirical observation, not on any underlying mathematical theory. They are often discovered through gut feel or what I've called the "ins": instinct, intuition, insight, inspiration, innovation, invention, etc. Their justification is therefore based on the fact that they simply work in practice. Although we preach "management by fact" it is important to also acknowledge that in many instances, and through mechanisms that we do not even understand, some people are able to see through process complexity and identify the underlying drivers.
Design of Experiments and the Taguchi Method
Another way to determine the influence coefficients would be to vary each of the independent measures over an appropriate range while holding all of the others constant and observing its effect on the dependent measure. By doing this we could also identify their range of independence. But in many instances, the number of required experiments would be impractical in both time and cost. Fortunately mathematicians have devised efficient experimental sequences in which we can vary more than one independent measure at the same time. The first to do this was Euler (1783) in what are called Latin Squares. Today such experimental schemes go under the name "Design of Experiments" or DOE. DOE is a popular tool used by six-sigma practitioners, and facility with it is usually a prerequisite for black-belt certification.
Genichi Taguchi has attempted to demystify DOE by creating a somewhat simplified procedure, that although not as mathematically rigorous usually gives an adequate answer. In doing so, he followed the example set by Walter Shewhart in his pioneering efforts (c. 1930) to bring statistical techniques to the shop floor environment.
Most statistical software packages now include a DOE and/or Taguchi Method capability (see for example Minitab, which is used in several six sigma initiatives). However, even with current software support, their use is beyond the capability of most improvement team members and requires expert assistance (e.g. staff statisticians or six sigma black belts). Fortunately, the vast majority of improvement efforts do not require this level of analysis in order to uncover the relevant independent measures.
In a process simulation, we attempt to dynamically reproduce its important characteristics in a computer model. By "running" the model, we can understand the complex interrelationships that exist within the process and test the effect of changes. Simple simulations are often done using spreadsheets such as Microsoft Excel or Lotus 123. For example, the columns in the spreadsheet might represent sequential times (e.g. months or quarters) while the formulas for each period's cells depend on several results calculated for an earlier period. Many software packages have specialized structures that make them particularly suitable for certain types of process simulations.
Flowcharting is an essential step in process improvement. Several of the current flowcharting software packages also include a simulation capability (I use Scitor Process) that is very helpful in finding internal leverage points, particularly when there are complex process flows and/or random variation is important.
A biotech company's product involved a new medical procedure that required special approval from the patient's insurance company for reimbursement. Long average approval times were having a serious adverse financial impact on the company. Furthermore, the variation (standard deviation) in approval times was also unacceptably high. What could they do to improve this result metric? There were many theories as to the root cause, most of which involved problems in someone else's area. The process flow was complicated by many alternate paths and frequent "resubmittal loops." A simulation of the process (using the Monte Carlo method), based on probable paths at each process node explained both the average and variation in approval time and pointed directly at the independent measures whose improvement would have the greatest impact.
For complex processes that contain time lags as well as subjective variables, System Dynamics modeling can also be very valuable (here I use Vensim). System Dynamics modeling has the advantage that it can easily accommodate both non-linearity and interdependencies, although its successful use does take considerable modeling skill.
To successfully compete in a new market segment, an electronics company needed major improvements in its delivery performance. Stratification of late shipment data showed that it was significantly higher the last week of the quarter. Again, there were many theories as to why. A system dynamics model of the entire order fulfillment process (order receipt to payment by customer) uncovered the answer and it was closely related to a similar phenomenon known as the end-of-quarter revenue "hockey stick."
Shipment linearity implies that with constant revenues, one-thirteenth of the quarterly total accumulates each week. In many organizations, there is a shortfall and the revenue falls below this linear goal. Miraculously, in the last week or two of the quarter, a few heroes appear and through their superhuman efforts the target is achieved and they are appropriately rewarded. The shape of the resulting weekly cumulative revenue curve is much like that of a hockey stick, whence its name.
The model explained what was happening. The added revenue at the end of the quarter came from early shipments of large dollar orders not due until the first few weeks of the following quarter. With limited capacity, this was at the expense of many small orders due in that hectic end-of-quarter period. Even worse, once started, this practice triggered a perpetual cycle where only small quantity unfilled orders were due for shipment at the start of the next quarter thus creating that initial revenue shortfall. The solution: just as the cycle was started by a one-time action, it needed to be ended the same way -- just stop doing it! Unfortunately, this results in a temporary sales shortfall that only goes unnoticed if it is hidden by rapid revenue growth. By phasing the practice out over several quarters, the adverse revenue impact was minimized.
Without the use of a simulation model, it would have been difficult to identify either the root cause or a palatable corrective action plan.
Choosing the Scorecard Metric
If improving a particular results measure is a strategic goal, then improvement efforts should be focused on the process measures that will have the highest impact on its improvement. They are usually the process measures with the largest influence coefficients. What does that imply about choosing scorecard metrics?
Most scorecards that I've seen are heavily populated with results metrics. No doubt this results from the all too common management attitude: "I don't care how you do it, just do it!" I strongly believe that ALL scorecard metrics must be directly actionable by their owner. Therefore, it's the underlying process metrics, not the results metrics that belong on a scorecard. If the improvement goals for the process metrics are achieved, than we can be assured that the desired results will follow, assuming we have identified these drivers correctly.
For example, dieters often tend to focus on their body weight (a results metric) rather than its independent measures: exercise along with calorie, protein, fat, and carbohydrate consumption. Nutritionists now believe that successful diets involve lifestyle (aka process) changes that act on these independent measures. Get them right and over time you will achieve and sustain your weight goal. I wonder to what extent this results focus explains the statistic that 95% of dieters fail to maintain their weight loss.
I would argue that results metrics only belong on a scorecard when their associated process metrics are on two or more subordinate scorecards. In this case, the job of the owner of the results metric is not its improvement, but sponsorship of the subordinate scorecards. That sponsorship includes guidance, monitoring and diagnosis, organizational troubleshooting, resourcing, communicating, etc. for the individuals and teams responsible for the subordinate scorecards. There is an important place for results measures, but it is mainly in the detection step in process control, not improvement.
The Japanese have a saying "Focus on process, not on results." In no case is this truer than in the selection of scorecard metrics. The key to linking strategy to action is not the balanced scorecard itself; it is this underlying process focus.
1 The influence coefficients are given by the partial derivative of fj with respect to xi.
2 See for example: Shoji Shiba, Alan Graham, and David Walden, "A New American TQM: Four Practical Revolutions in Management" Productivity Press Inc., January 1993, ISBN: 1563270323, pg. 221.
Summary and Conclusions
In this e-paper, I have argued that the BSC must be viewed as part of an organization's on-going strategic planning system. I've described that process using a 9-step model that integrates many of the contemporary elements of effective strategic planning including strategic focus and intent, critical success factors, core competencies, stakeholder satisfaction (and their associated result metrics of loyalty and retention), process improvement, performance measurement, and assessment.
The linchpin in this process is a focus on stakeholders and their unsatisfied requirements. Understanding who the strategically determined stakeholders are and the important opportunities for improving their satisfaction forces the process to move from ubiquitous statements (e.g. "... to be the market leader") to tangible and assignable actions directed toward that end.
these concepts into a single comprehensive process in the environment of
increasing complexity and chaos requires the use of unfamiliar tools.
Although admittedly tedious, their use is not difficult given proper
facilitation and the use of readily available automation support tools.
The benefits of this approach can be significant. Done correctly it draws on the knowledge that is widely dispersed throughout the organization and integrates that knowledge to arrive at a strategy that best utilizes all of the organization's assets: physical, human, and intangible. Although the allure of a quick solution is ever present, it takes time to draw together existing knowledge and generate new knowledge where necessary. But what can be more important to an organization's survival than a well-conceived strategy and an associated focused action plan that enrolls the entire organization in the required course of action that will assure its achievement?