Uses and Abuses of Dashboards

Mark Morgan and Rita E. Numerof, for HealthLeaders News, October 2, 2007

Dashboards provide an at-a-glance overview of progress against goals, and have gained wide acceptance as a tool for managing organizational performance--with good reason. When they provide a valid, holistic view of progress, they focus attention on critical operational strengths and weaknesses, allowing corrective action to be taken where it is needed, when it is needed.

Dashboards themselves can be technologically simple paper-and-pencil affairs capturing data regarding an individual salesperson's ability to move prospects through the sales pipeline. They can also be hugely complex pieces of IT infrastructure pulling data from every part of a global organization and displaying it in real time for the benefit of top executives. But whether simple or complex, and whether intended to guide performance management discussions or strategic resource allocation, the concept is the same. The dashboard should summarize metrics that, taken as a whole, produce a valid and reliable picture of how well current activity is producing progress toward strategic goals.

But dashboards frequently mislead for two primary reasons. First, the metrics on which the dashboard is based may be invalid--bearing little real relationship to strategic ends. Second, the metrics may create "local optima"--they may drive people, or entire units, to engage in activity that maximizes their own benefit, even if it produces suboptimal outcomes for the organization as a whole.

Invalid Metrics
Invalid metrics often measure activity rather than results, which can result in behaviors that provide no value whatsoever. Metrics designed to ensure the completion of paperwork without providing feedback regarding the quality of the data demonstrate this flaw. Reams of data are produced and transmitted, and all of it is useless for the operational analysis it is intended to enable.

But rewarding activity is not the only way a metric can produce behavior that is counter to the organization's goals. Sometimes a metric drives genuinely productive behavior that isn't aligned with the organization's strategy. Let's say that a hospital wants to establish a reputation as a patient-friendly place, with the long-term goal of becoming a provider of choice--able to command a premium on that basis and thereby boost profitability. Ideally, dashboard metrics should reflect that strategy. But in our hypothetical hospital they instead reflect a short-term need for cost control, driving people through admissions as efficiently as possible. That gets translated by hospital staff as "business only, no chit-chat allowed". The metric drives efficiency and is aligned with the short-term goal of operating profitably. But it is out of alignment with the strategic direction of the organization: where's the friendly face? The dashboard flashes green when it should be red.

Local Optima
The other way that dashboards commonly cause problems is by producing "local optima", ways that an individual or unit can make itself look good at the expense of other individuals or groups within the organization, and hence at the expense of the organization as a whole. This happens when optimal outcomes require collaboration, but the metrics and associated incentives drive silo-like behaviors.

This is all too common. We've seen people in a bank refuse to introduce their clients to other service providers within the same bank because there was some chance, however slight, that the client might be alienated by the new person--which might drive the client to take his or her business elsewhere. We've also seen hospital employees fail to make required entries in patient records databases because they have more important things to do (according to their metrics), and they don't need the data themselves.

In both cases it's likely that the dashboard is red somewhere--but not where the problem lies. The bankers who should have access to a rich vein of potential new clients (but don't) and the care coordinators who should have immediate access to patient data (but have to dig it up piecemeal instead) both look bad, but the problem doesn't lie with them. It's their green-blinking compatriots that are the problem.

Building a better dashboard
Building a better dashboard is not just a matter of providing more performance data, more quickly, and in a more compact and efficient format. To build a good dashboard, you need to start with the right data--metrics that are inextricably linked to the strategic goals of the organization. That's a tall order in many organizations, because the strategic goals aren't clearly defined or communicated. So the first priority is this: seek clarity. What outcomes is the dashboard supposed to drive for the organization? What implications does that have for unit-level metrics? What does that mean for individuals within those units? Cascading of goals and metrics is an important part of developing a working dashboard, but the more nebulous work of cascading the implications of strategy needs to come first.

The need to focus on the right data, defined in alignment with strategic goals, also has implications for that most beloved of modern innovations--the "balanced scorecard". A kind of dashboard, the scorecard uses metrics in an effort to align individual, unit, or organizational performance. For some reason, however, balanced scorecard users are often profligate in their use of metrics. Instead of focusing on one or two really critical measures that truly reflect progress toward strategic goals, dozens of metrics, sometimes fifty or more, are defined--each contributing very little to an overall score. "Measure everything that's easy to measure" appears to be the maxim, which means focusing on raw activity.

As a result of this excessive focus on tactical activities the overriding objective--contributing value--often gets lost. Anything not included on the scorecard (which often means less easily measured collaborative and planning activities) is viewed as a distraction. Alignment with strategic goals is lost.

Each person and each unit has a few key ways that it contributes to the strategic objectives of the organization. Finding a small set of metrics that measure this contribution, and that encourage optimal cooperation by measuring the higher-level outcomes that depend on close collaboration, helps to bring activity into alignment with strategy.

The second set of questions to be confronted is this: what metrics are ideal, what metrics are practical, and how do we bridge the gap? The answer is not always simple. Complications arise because, even within a role, expectations may vary depending on the circumstances. Someone in a "training-phase" is expected to spend time learning how to use the tools of the trade, follow procedures, and so on--it's not that they aren't expected to be as productive as more experienced colleagues, it's that the things that constitute productivity are truly different. A single dashboard that measures both training-phase activities and activities expected of someone who has matured into the role pushes both groups into activities that are inappropriate. Two distinct sets of metrics are needed.

Likewise, metrics can be constructed to reflect collaborative imperatives. In some cases, this means moving the metric up a level--if various staff aren't collaborating well on the fifth floor, a measurement at the level of the floor will tend to reflect that even if metrics at the individual level don't. In other cases, it may make sense to create a metric that measures collaboration directly--allowing the care coordinator to indicate that patient records were not available as they should have been, and have that reflect back on the person or group that should have made them available.

Ensuring that dashboards reflect the strategic needs of the organization takes skill, cooperation (especially at the executive level), and effort. But the promise of a dashboard system is that it will create strategic alignment across the organization, detect any deviations from that alignment, and allow for prompt evaluation and corrective action. Like a pair of glasses for management, it should bring operations into clear focus. Imperfect lenses produce organizational headaches.

---

Mark Morgan, M.S.(R)., is a senior analyst, and Rita E. Numerof, Ph.D., is president of Numerof & Associates, Inc., a strategic management consulting firm located in St. Louis. For more information, contact NAI at 314-997-1587, info@nai-consulting.com, or visit http://nai-consulting.com/.