Balanced Scorecard Best Practices: Understanding Leading Measures Author: Bill Barberg One of the concepts in the Balanced Scorecard methodology that appeals to many executives or other business decision-makers is the idea of having “leading measures.” After all, nobody wants to be compared to a driver who can only navigate by looking through the rearview mirror. When Dr. Robert Kaplan and David Norton introduced the Balanced Scorecard (BSC) over a decade ago, part of the “balance” that they introduced was balancing the traditional financial measure, which they characterized as lagging measures, with measures that gave a better indication of likely future performance—leading measures. The Context: Understanding Balanced Scorecard Terminology Before going into details on the concept of leading measures, it is helpful to understand a few key Balanced Scorecard terms. (After all, one major benefit that Kaplan and Norton have touted is the establishment of a standard set of terms and definitions that can allow people to communicate more effectively with regard to strategy execution and performance management.) Some of the most important terms are perspectives, objectives, measures, targets, initiatives, and cascading. The Balanced Scorecard methodology stresses that objectives and measures from multiple perspectives should all be considered. The classic perspectives for for-profit businesses are Financial, Customer, Internal Operations/Processes, and Learning and Growth (which focuses on human capital, technology and organizational culture—the intangible assets that create value). By looking carefully at all four perspectives, organizations can focus on both the causal drivers of performance and the outcomes. In the Balanced Scorecard, the strategic objectives (often just referred to as objectives) are the stars of the show. Objectives often consist of a verb-adjective-noun phrase. For example, an objective may be something like “Grow International Sales” or “Build Deep Client Partnerships.” These objectives should be linked in cause and effect chains that cross the multiple scorecard perspectives—graphically depicted in what has become known as a strategy map. The following diagram shows objectives linked in cause and effect chains that are part of a strategy map for a software company based in the U.S. that wants to execute a strategy of growing international sales. 1 Figure 1 - Part of a Strategy Map for a Software Company The measures are used to monitor progress toward accomplishing the objectives. Measures don’t stand alone. They belong to specific strategic objectives. In the prior example, the measure for Grow International Sales could be “$ in International Sales” and the measure for Local Sales & Support could be the “Number of Trained Resellers outside the U.S.” Targets are established for the measures, along with the logic for color bands so reports can easily show if a particular measure is red, yellow or green (or perhaps some other color, depending on the logic that is used.) There are often multiple targets for different time horizons. For example, the target for international sales may be $10 million for the third quarter and $12 million for the fourth quarter of the current year. For the fourth quarter, red may be defined as less than $11 million and yellow as $11 million up to $12 million in international sales for that quarter. Strategic initiatives (often just called initiatives) are the specific projects, outside the normal activities of the organization, that are undertaken specifically to help accomplish the strategy or close the gap between a measure and the target. Initiatives almost always have start dates and end dates and a focus on accomplishing something that will support the accomplishment of the particular strategic objective as measured by that objective’s measure. For example, in the case of the software company wanting to grow internationally, an initiative might be something like, “Develop a list of possible 2 resellers outside the U.S.” In contrast, something like “Pursue partnering opportunities” is too vague to be a good initiative, and it is not likely to have a start and end date. (That could be an objective— perhaps on a supporting scorecard for the business development department—and it would then need to have a measure associated with it.) On the other hand, the details surrounding each of the on- going sales and marketing activities to recruit new partners are probably best treated as tactical or operational information rather than a series of strategic initiatives. If too many operational tasks get treated as strategic initiatives, the clutter can diffuse the focus on the critical strategic initiatives. The concept of cascading involves taking the top level organizational strategy, as clarified by the objectives on the strategy map, and rolling that down to lower levels of the organization, such as business units, divisions, departments or individuals. The process of cascading is not just a matter of pushing the same measures down to other parts of the organization. Instead, the same type of cause and effect thinking that is used to create the strategy map should be used to link the objectives from lower parts of the organization to the higher level objectives on the top level scorecard. In the above software company example, the business development department may have objectives for participating in foreign trade shows or for having meetings with the executives of potential reseller partners in foreign countries. The IT department may have objectives relating to the infrastructure needed as prerequisites for deploying a secure partner extranet. Those objectives, if successfully accomplished, would advance the execution of this company’s strategy by enabling improved international channel support, and therefore contribute to the financial objectives of growing international sales and generating sustained profits. The cascaded objectives should have tight cause and effect alignment with the top-level objectives, but they are not the same objectives, and they are not likely to have the same measures. If we think of the causal links on the strategy map as two-dimensional (2-D), because they can be represented on a 2-D sheet of paper, then we can think of the cascaded cause & effect relationships as three-dimensional (3-D). 3 Figure 2 - Cascading a Strategy Map Leading Measures Now that we’ve defined some context and terminology, let’s focus on the concept of leading measures. Leading measures are not a crystal ball. Nor should they be confused with a forecast. Somebody’s prediction about next month’s sales is not a leading indicator. It is just a prediction about the future value of a lagging indicator. In practice, there are two types of measures that are often characterized as leading indicators. The first is a leading indicator with regard to accomplishing a specific objective. For example, if a manufacturer of electronic components has an objective of reducing turnover in their customer support staff, then a leading measure for that objective might be the percent of employees in the customer support department who indicate that they would recommend to a friend that they should try to get a job in this organization’s customer support department. If people are recommending that their friends try to get a job similar to their own, it is a good indicator that the employee is happy with the job and company and is not about to jump ship. Therefore, this would be a good leading indicator for the objective of reducing staff turnover. A lagging indicator for this objective might be the actual turnover in the customer support department. As it relates to that specific objective, actual turnover is a measure of what has already occurred, so it is looking in the rearview mirror. However, if you step back and look at the full cause and effect chain, the objective of reducing turnover in the customer support department is an important causal driver of improving the quality of service and expertise that they provide to their customers. The strategy is to win customer loyalty and improve profit margins by competing on superior service instead of price. The strategy map visually represents key causal relationships in the strategy. 4 Figure 3 - Part of a Strategy Map for a Manufacturing Company If the hypothesis is true that reducing staff turnover will improve the quality of service and the degree of expertise that they can offer their clients, and that offering that improved support will result in greater customer loyalty and higher profit margins, then it is perfectly reasonable to think of actual turnover in the customer service department as a leading measure with regard to the outcome objectives of improving customer loyalty, sales, and profit margins. Therefore, depending on which context we are looking at, actual employee turnover could be either a lagging indicator (with regard to the objective it is associated with) or a leading indicator (with regard to the whole cause and effect chain). It would be a mistake to waste too much time debating whether any one measure is really a leading indicator or a lagging indicator. That misses the underlying principles. An important principle of the Balanced Scorecard methodology is to get people to think in terms of cause and effect relationships, and to pay close attention to both the cause and the effect. Another important principle is to focus on the critical few objectives and measures, and to have the appropriate people in each part of the organization looking at the objectives and measures that they can influence or that provide them with valuable insights about their situation and their contribution to the higher-level objectives. An implication of these principles is that not everyone needs to be looking at the same measure. In this specific example, the desire to keep the measures to manageable level 5 (<30 for most scorecards) may dictate that the top level corporate scorecard just include the measure of staff turnover in the customer support department. That would be a leading indicator in the context of the cause and effect linkages. However, as the scorecard cascades down to the next levels, there may be supporting objectives (or perhaps a repeat of the same objective) on both the Human Resources scorecard and on the Customer Support department scorecards. On these scorecards, which have a much narrower scope, it would be more appropriate to include both measures that relate to the staff turnover in the customer support department. There is a lot of truth in the saying that “You get what you measure” or “What gets measured gets done.” When organizations focus on the leading indicators, they are exerting a powerful influence on accomplishing the things that will drive the successful execution of the strategy. The significance of this should not be underestimated. Research has shown that only between 10% and 30% of the strategies are successfully executed in today’s organizations. Most organizations put the emphasis on the lagging indicators like sales revenue or profit. Focus on those measures can certainly put pressure on people to hit their targets, but if there is not a clear strategy and an understanding of the causal drivers, that pressure can lead to people taking short-term, conflicting, and counter-productive steps, often undermining the long-term success of the organizations. For example, organizations often suffer when desperate salespeople resort to drastic measures to try to hit their sales quota at the end of the quarter—setting up a margin-killing environment where prospects delay purchases because they know they have great negotiating power in the last days of each quarter. Or, we’ve all probably seen organizations that lay off some of their most valuable staff because they are pressured to meet some financial ratio, only to hire these people back at much higher consulting rates because they need the skills the people offer. More common is the situation where multitudes of people in an organization are working hard to achieve their financial goals (the lagging measures of their success) with a serious lack of alignment and focus. Even if the leaders of an organization have a strategy, without the clearly defined objectives and appropriate leading measures, the odds of successfully executing that strategy are low. Without strategically focused measures, people’s efforts are often diffused, sub-optimized and disjointed. As evidence of how common this alignment problem is, a recent survey by Bain & Co. cited in Optimize magazine (Oct. 2004), revealed that while 70% of senior executives believe IT is highly relevant to enabling their companies to grow, 60% complain that IT is inhibiting key activities that foster growth. In contrast, organizations that put the time and effort to determine the critical non-financial drivers of their success, using tools like strategy maps, end up performing much better on the long-term financial results. Wharton Professors Christopher Ittner and David Larcker summarized their research on the performance management systems for 157 companies in a November 2003 article in Harvard Business Review. Their research found that 23% of the organizations had done extensive causal 6 modeling to identify non-financial measures that were linked to their strategy. Those organizations substantially outperformed their peers who had not done the same causal modeling, showing, on average, a 2.9% higher Return on Assets and a 5.14% higher Return on Equity. Ittner and Larcker’s research showed that there is a fundamental difference between merely selecting a “balanced” mix of measures and identifying the true strategic leading measures that help drive strategy execution and long-term financial success. Leading Measures Rarely Add Up Well chosen leading measures should be associated with objectives that are aligned based on cause and effect relationships (either 2-D relationships on a single strategy map, or 3-D causal relationships on cascaded maps or scorecards). The objectives and measures on these causal chains are typically different from one level to the next, which means that you usually cannot mathematically aggregate measures from one level to the next. If we consider the Marketing Department of the software company, creating foreign language marketing materials should help drive increased international sales, but you can’t add up the number of languages that brochures are available in to get the international sales figure. Similarly, if the Business Development department increases its participation in foreign trade shows and in meeting with the executives of potential resellers, those leading measures should drive improvements in the number of trained resellers outside the U.S., which is the measure associated with the objective of providing customers with local sales and support—which is a causal driver for increasing international sales. The relationships between these leading indicators are often fundamentally different from commonly used lagging measures. The “cascading” of many lagging measures is often a simple matter of aggregation—not cause and effect chains. For example, sales revenue is typically aggregated from the details of the transactions. After it is rolled up, it is common to drill down in a variety of ways to understand the details. Sales revenue can be sliced by divisions, districts, territory, product groupings, promotional campaign, and numerous other ways if the data is properly collected and aggregated. The same is true for many financial measures, like expenses or profit. When working to deliver useful performance metrics for many of these lagging measures, one of the biggest challenges is to get people using the same definitions. For example, if we are measuring sales revenue, it is important to have consistency with regard to when revenue is recognized and what is included. When looking at things like gross margin contribution, it is important to have consistent definitions and well-defined aggregation logic if it is going to be possible to do the roll-ups, drill-downs, and other manipulations. Many powerful software packages, data warehouse techniques, and other IT tools have emerged to address these issues. OLAP tools allows super-fast slicing, dicing and drill-down. Budgeting tools allow multitudes of inputs to be managed, compared, analyzed and massaged in a multitude of ways, based on the carefully designed logic for categorization and summarization. 7 Unfortunately, the strengths of many of these applications become problematic when used in the realm of leading indicators. With leading indicators, it is generally NOT a priority to be using the same definition of measures at the various cascaded levels. Most of the leading measures are different from level to level anyway, as illustrated in the earlier examples. And the type of drill-down required is not primarily based on mathematical aggregations, so the multi-dimensional slice and dice functionality is often not appropriate. The problems come when the supposedly “leading” measures get selected based on fitting into software functionality that is optimized for certain types of lagging measures. Rather than selecting measures that reflect progress in achieving critical strategic objectives that are linked by cause and effect relationships, organizations often select measures based on the fact that the same measure can be used at each level of the cascading process. These “least common denominator” measures rarely have strong links to strategy. They are usually so generic that they provide little direction to help create strategic alignment. The Conflict Between Most BSC Software and BSC Best Practices for Leading Measures This sets up a conflict between the people who are trying to follow Kaplan and Norton’s principles (and the solid empirical research that supports the value of selecting leading indicators based on causal drivers), and the highly-appealing functionality for drilling-down, slicing, and dicing you performance measures. Management consultants who teach and train on the Balanced Scorecard methodology have often taught that it is best to NOT introduce software into the scorecard design process because it inappropriately steers the measure selection toward measures that have lots of detailed data and/or aggregate well, rather than those that are the most powerful causal drivers. But without the help of technology, organizations that take this path often get frustrated and bogged down with all the manual effort and the building of spreadsheets, or other “manual” data management tools. Meanwhile, software vendors battle for market share by showing off their functionality as balanced scorecard tools. This typically involves demonstrations of various dashboard displays, executive briefing books, color-coded thresholds, gauges and displaying metrics from four perspectives. There is increasingly some sort of strategy-mapping functionality, but rarely is there the ability to show and navigate on the cascaded “3-D” cause and effect chains that are one of the most powerful best practices in the Balanced Scorecard methodology. One downside of most of these software packages is the fact that they only work well for data that fits certain characteristics—a standard set of measures that are cascaded from one scorecard to the next, logical hierarchies for aggregation, and consistent base data that is suitable for aggregation. These are not the characteristics of most good leading indicators. But when organizations deploy most software packages, the appeal of the aggregations (or the efforts to get the software features to function so purchase of the software can be justified) is often stronger than the desire to have measures with causal relationships. Therefore, many organizations that try to deploy software as part of their Balanced Scorecard fall far short of the 8 potential benefits they could achieve by using the “best practice” principles for selecting leading measures. Insightformation, which had a consulting background in Business Intelligence and the Balanced Scorecard, saw this conflict between software functionality and Balanced Scorecard best practices as an opportunity. We set out to create a different type of software that would support the principles of the Balanced Scorecard methodology, especially with regard to cause and effect relationships. We recognized the value of Business Intelligence functionality (like slicing, dicing, and drilling down on data), but we felt those capabilities were best positioned as an analytical drill-down option for certain measures—not as the primary architecture for the scorecard functionality. Our rationale for how we recommend leveraging BI in a BSC solution is further addressed in my earlier article, Business Intelligence and Balanced Scorecards: Different Paradigms. By putting an emphasis on strategic objectives and cause and effect relationships, our goal was to have a Balanced Scorecard automation option that would encourage the selection of leading measures that could best drive strategic alignment and breakthrough performance. If you study the organizations who achieved breakthrough results based on the Balanced Scorecard, you will often find a few leading measures that played a particularly powerful role in driving strategic change. Defining those critical leading measures is not easy, but the effort can pay a big reward in focus, alignment and successful strategy execution. About the Author Bill Barberg is the President and Founder of Insightformation, Inc. Founded in 1991, Insightformation has been a pioneer in helping organizations leverage information and technology to gain insights and improve performance. In 2001, Insightformation was hired by Microsoft to create the Microsoft Balanced Scorecard Framework (BSCF). In 2002, they developed InsightVision, a software application based on the Microsoft BSCF. In August 2003, InsightVision was awarded a Grand Prize in the Microsoft Office Partner Solution Builder Contest. Contact Bill at 763-521-4599 x13 or bill.barberg@insightformation.com. 9 . achieve by using the best practice” principles for selecting leading measures. Insightformation, which had a consulting background in Business Intelligence. 1991, Insightformation has been a pioneer in helping organizations leverage information and technology to gain insights and improve performance. In 2001, Insightformation