Deployment of the Balanced Scorecard Measurement System
The Balanced Scorecard text by Kaplan and Norton describes in general terms the basic ideas of the balanced scorecard concept, its great advantages over past approaches to strategic management, and a general outline of how to develop and deploy such a system. However, there are many issues involved in making a good fit to a specific organization, and in estimating and minimizing the cost and time of deployment.
Change Management Comes First
Change management is always going to be a major challenge in deploying the balanced scorecard system, because it must touch every significant activity in the organization. John Kotter in Leading Change provides a description of the conditions that should prevail in order to ensure success. The most important of these is a sense of urgency in the senior management of the organization. Does your agency's management have a sense of urgency, or complacency? How can you instill greater urgency in them? This effort should not be underestimated, because success depends on their consistent support and understanding of the deployment.
Since the balanced scorecard process affects everyone in the organization, it has to confront the organization's culture. Many government agencies have a culture of complacency and even cynicism, due to the downsizing that has been prevalent in recent years. As young people are hired, it is critical to prevent these attitudes from perpetuating themselves. The balanced scorecard, by providing greater openness and visibility of an agency's mission and strategy, can give rank-and-file employees a better sense of what their jobs are about, and empower them to improve how that work is done. Strategic goals, balanced scorecard plans and improvement plans should be posted on an intranet web site for all employees to see. Broad participation in planning should be encouraged.
Deploying the Balanced Scorecard as a Pilot Project
Kaplan and Norton provide an outline of how to deploy the BSC. Basically, a core team is formed to develop the detailed plan and deploy it. To reduce cost and risk, it is usually wise to initially deploy pilot projects in a limited part of the organization. However, all deployment efforts should be initiated at the headquarters level. Strategic goals have to come from the top. Also, some strategic initiatives to reach these goals should be identified there. Some metrics should be standardized at the top level in order that measurements from different departments will be comparable for benchmarking. At the departmental or pilot project level, specific goals, initiatives, desired outcomes, metrics, targets, and annual milestones can be defined. The constitutes the strategic plan plus performance plan in accordance with Government Performance and Results Act.
Balanced Scorecard Measurement Processes
There are two sets of more-or-less continuous data flows required in the BSC system:
1. Downward information flow
Line managers at the Directorate, Department, and Branch levels define goals, desired outcomes, initiatives, metrics, targets, and schedules. The goals, metrics, targets and schedules are aligned with those specified in the top-level strategic plan and balanced scorecard performance plan. Some ofhese parameters may have to be translated from general to more mission-specific to apply to the work being performed at each organizational level. Specific desired outcomes and initiatives to attain them are developed by these managers, and the metrics, targets and schedules are then developed. This results in a hierarchical set of balanced scorecards that are pertinent at each organizational level. It also has the effect of giving the organizational stakeholders a meaningful role in performance evaluation and strategic management.
2. Upward information flow
Define collection methods for each of the BSC metrics. This is the most expensive, labor-intensive, aspect of the BSC system, and has the most impact on the rank-and-file employees. I therefore recommend that for survey-based measurements, surveys should be carefully and professionally designed to minimize complexity and impact on the employees' time. Aggregate metrics data by organizational code. Create reports at each line management layer. These will provide feedback to the managers on metrics that are pertinent to their own strategic interests. Aggregate at each level (without filtering or loss of data). At the senior management layer, combine the metrics data into the 'vital few' needed to give an overall picture of the agency's balanced scorecard, on a continuous basis.
These two information flows are illustrated in the following diagram, using a notional organization chart:
In this figure, the top-level strategic goals, metrics and targets are in red, and they flow downward from the headquarters strategic planning office. The balanced scorecard measurements, in blue, are collected starting at the bottom (branch level) and flow upward. (This is analogous to the circulation of blood from arteries to veins.)
At each level of the organizational hierarchy, the data are aggregated across the lower levels. Aggregation serves to reduce information overload. Periodically, measurements are collected, aggregated and analyzed at each management level. These data provide the managers with knowledge of strategic performance within their own part of the organization. In other words, the performance evaluations are not only for the top-level managers. Each manager, at each level, benefits by seeing the same metrics as they apply to his or her own area of responsibility. This is an incentive to support the balanced scorecard: we are measuring ourselves -- and we get to use the data we collect for our own purposes.
Use Information Technology to Support the System
Once the metrics and data collection procedures have been defined, an information infrastructure can help greatly in managing the data flows. A database-backed web intranet is recommended both for data collection and data reporting. This can be developed at relatively low cost using mostly COTS products, or components of it can be purchased. Web technology can support survey data collection and data reporting fairly easily. It also can enable features such as workflow process monitoring of cycle times, activity-based costing, and an 'anonymous channel' to upper management. Agencies such as DoD components and research labs often have the skills to build at least part of the system in-house. Information technology papers can be found here.
We should not leave the reader with the impression that deployment of the balanced scorecard in the government is easy or straightforward. In addition to the change management challenges mentioned above, there are several other issues that have not been sufficiently addressed in other literature. These issues confront implementers of the balanced scorecard in many agencies, and deserve continued serious study. Some of the major ones are described here.
1. Organization: In a matrix organization, most of the mission-oriented work is done in program/project teams, not according to the branch-department-division hierarchy. So should mission-oriented metrics be defined by the program managers rather than line managers?
Top-level metrics tend to be very generic: core knowledge, experience, capability, performance, realism, facility maintenance, etc. There are no metrics to describe the quality of research, the effectiveness of products and services, or the long-term outcomes of the mission. Rather, these are inferred from the level of customer satisfaction. If customer satisfaction and perception is the focus of the BSC, then it will be sufficient to define the metrics within the line management structure.
On the other hand, if headquarters desires more pertinent metrics on our mission-oriented quality, success, and outcomes, then it will be necessary for program managers to define specific metrics within their technical areas. Advantages are that they provide leading indicators for future customer satisfaction, they create an incentive for improvement within the programs, and they are probably being tracked informally already. The disadvantage is that unlike the generic metrics, these will be different for each program, harder to define, and probably not commensurate across programs to permit benchmarking. For example, how do we compare progress in public health care with progress in improving cost-effectiveness of a bridge design?
2. Cost. There are vendors that now provide SOME of the software and methodology needed to support the balanced scorecard. Several vendors offer systems that handle the planning, data aggregation and reporting/display functions. However, such systems, marketed primarily to Fortune 100 companies, are expensive, and organizations may need to consider an in-house build option. Moreover, such systems cannot COLLECT the data, which is a labor-intensive need. Cost minimization is an important factor in survey design; it may be cost-effective to bring in a consultant to design and/or conduct the surveys. This issue is discussed further in the references cited in survey methods. At any rate, a total cost calculation should be done to identify the best implementation approach.
3. Fear. At the branch level, it would be convenient to link the individual employee performance evaluations to the balanced scorecard. But is this a good idea? If too much emphasis is placed on performance of individual employees, the BSC could change into a warmed-over version of "Management By Objectives", which failed back in the 1970's. It might scare employees, and it would take the incentive and onus off managers to take responsibility for strategic management, the intent of the BSC approach.
On the other hand, since employee performance is to be evaluated in any case, these data should not be ignored. Moreover, there is a stated intention for agencies to set goals and objectives for personnel performance. Thus, tying the BSC process to the personnel evaluation process should help to align the organization strategically. However, in a public-sector organization, fear must be balanced by an incentive -- compensation, promotion, increased bugetary authority or other tangible benefits. In the government or nonprofit arena, fear unbalanced by real incentives will kill the balanced scorecard before it gets off the ground.
A suggestion to deal with this issue has been to stipulate that the BSC data are only to be used in the aggregate, not for individual evaluations. The collective performance of branches, programs and/or teams is the concern of the senior managers, not the performance of individuals. If this is clearly understood by all, then the fear involved in employee survey data collection or use of demo data should be minimized. However, this still leaves the employees and many managers asking, "what's in it for me?"
4. Objectivity. Charles Savage (Fifth Generation Management) notes that
"As information passes up and down the organization, there is a strong tendency to manage and massage it. A function is often more interested in making itself look good than in telling things the way they are. In too many companies, even before the computer, we have had managed information systems. The computer has not changed this pattern of interaction between levels. Usually MIS attempts to mirror the configuration of steep hierarchies, so they suck up information from lower levels, summarize it, and provide top management with executive decision support systems. These decisions are then communicated back down through the hierarchy to lower levels.... Is this where managed information systems are taking us? Will these systems be used to 'micromanage' the plants and their functions? .... if little trust exists between levels or between functions in steep hierarchies, there is going to be a greater interest in managing or massaging the information that is passed vertically between boxes. Is it possible that the wrong questions are being asked in the MIS arena? The more appropriate questions may deal with leveraging knowledge rather than managing information. The challenge is to see the interrelationship of multiple patterns. Unfortunately, traditional managed information systems tend to remove all the interesting rough edges and sanitize the information so the important patterns do not show up." (pp. 107-8).
How can we create an objective data collection system that minimizes these problems Savage identifies?
This article has pointed out the concepts and data management efforts that are involved in balanced scorecard deployment. It is clear that modern information technology systems can support this work. However, the big implementation problems are people problems: change of managers' perceptions, fear in the employee culture, and the limited incentives available within most governmental and nonprofit organizations to counterbalance that fear. But this is where we are now with the balanced scorecard in the federal government, and these are the same issues that any future management system will have to confront.
©1999 Paul Arveson