Creating Effective KPIs
One of the most common questions people ask about performance dashboards is, How do we define effective key performance indicators (KPIs)? The answer is important because KPIs govern how employees do their jobs.
Agents of Organizational Change
The adage "What gets measured, gets done" is true. KPIs focus employees' attention on the tasks and processes that executives deem most critical to the success of the business. KPIs are like levers that executives can pull to move the organization in new and different directions. In fact, among all the tools available to executives to change the organization and move it in a new direction, KPIs are perhaps the most powerful.
Subsequently, executives need to treat KPIs with respect. As powerful agents of change, KPIs can drive unparalleled improvements or plunge the organization into chaos and confusion. If the KPIs do not accurately translate the company's strategy and goals into concrete actions on a daily basis, the organization will flounder. Employees will work at cross purposes, impeding each other's progress and leaving everyone tired and frustrated with little to show for their efforts.
A trucking company, for example, that measures performance by the percentage of on-time shipments may drive hauling costs skyward because the KPI does nothing to discourage dispatchers from sending out half-empty trucks to meet their schedules. To keep costs in line, the company needs to add a second KPI that measures the percentage of unused cargo capacity in outgoing trucks, and it needs to revise the first KPI so it emphasizes meeting customer expectations for fast, reliable shipments rather than just on-time deliveries. This combination of KPIs gives dispatchers leeway to contact customers and renegotiate shipping schedules if they know the customer may be flexible.
The Zen of Development
Crafting sound KPIs is more of an art than a science. Although there are guidelines for creating effective KPIs (see sidebar), they do not guarantee success. A KPI team may spend months collecting requirements, standardizing definitions and rules, prioritizing KPIs and soliciting feedback - in short, following all the rules for solid KPI development - but still fail. In fact, the danger is that KPI teams will shoot for perfection and fall prey to analysis paralysis. In reality, KPI teams can only get 80 percent of the way to an effective set of KPIs; the last 20 percent comes from deploying the KPIs, seeing how they impact behavior and performance, and then adjusting them accordingly.
Metrics used in performance dashboards are typically called key performance indicators because they measure how well the organization or individual performs against predefined goals and targets. There are two major types of KPIs: leading and lagging indicators. Leading indicators measure activities that have a significant effect on future performance, whereas lagging indicators, such as most financial KPIs, measure the output of past activity.
Leading indicators are powerful measures to include in a performance dashboard but are sometimes difficult to define. They measure key drivers of business value and are harbingers of future outcomes. To do this, leading indicators either measure activity in its current state (i.e., number of sales meetings today) or in a future state (i.e., number of sales meetings scheduled for the next two weeks), the latter being more powerful because it gives individuals and their managers more time to influence the outcome.
For example, Quicken Loans identified two KPIs that correlate with the ability of mortgage consultants to meet daily sales quotas: the amount of time they spend on the phone with customers and the number of clients they speak with each day. Quicken Loans now displays these two current-state KPIs prominently on its operational dashboards. More importantly, however, it created a third KPI based on the previous two that projects whether mortgage consultants are on track to meet their daily quotas every 15 minutes. This future-state KPI, which is based on a simple statistical regression algorithm using data from the current-state KPIs, enables sales managers to identify which mortgage consultants they should assist during the next hour or so.
Challenges to Creating KPIs
Some of the challenges with creating effective KPIs include process nuances, activity measurement, accurate calculations and lifecycle management.
Process Nuances. The problem with many KPIs is that they do not accurately capture the nuances of a business process, making it difficult for the project team to figure out what data to capture and how to calculate it.
For example, executives at Direct Energy requested a repeat-call metric to track the efficiency of field service technicians, but it took the project team considerable time to clarify the meaning of the KPI. For example, field service technicians primarily repair home energy equipment, but they can also sell it. So, is a repeat call a bad thing if the technician also brings literature about replacement systems or makes a sale? Or, what if a homeowner only lets a technician make minor repairs to an aging system to save money but calls shortly afterward because the home's furnace broke down again?
Most business processes contain innumerable nuances that must be understood and built into the KPI if it is to have any validity, especially if the KPI is used as a basis for compensation. The worst-case scenario is when employees discover these nuances after the KPIs have been deployed, which stirs up a hornet's nest of trouble and wreaks havoc on both the performance management system and compensation policies.
Accurate Calculations. It is also difficult to create KPIs that accurately measure an activity. Sometimes, unforeseen variables influence measures. For example, a company may see a jump in worker productivity, but the increase is due more to an uptick in inflation than internal performance improvements. This is because the company calculates worker productivity by dividing revenues by the total number of workers it employs. Thus, a rise in the inflation rate artificially boosts revenues — the numerator in the KPI — and increases the worker productivity score even though workers did not become more efficient during this period.
Also, it is easy to create KPIs that do not accurately measure the intended objective. For example, many organizations struggle to find a KPI to measure employee satisfaction or dissatisfaction. Some use surveys, but often employees do not answer the questions honestly. Other companies use absenteeism as a sign of dissatisfaction, but these numbers are skewed significantly by employees who miss work to attend a funeral, care for a sick family member or stay home when day care is unavailable. Some experts suggest that a better KPI, although not a perfect one, might be the number of sick days taken because unhappy employees often take more sick days than satisfied employees.
Natural Lifecycle. It's also important to note that a KPI has a natural lifecycle. When first introduced, the KPI energizes the workforce, and performance improves. Over time, the KPI loses its impact and must be refreshed, revised or discarded. Thus, it is imperative that organizations continually review KPI usage.
Performance dashboard teams should automatically track KPI usage with system logs that capture the number of users and queries for each KPI in the system. The team should then present this information to the performance dashboard steering committee, which needs to decide what to do about underused KPIs. For example, Hewlett Packard Technology Solutions Group (TSG) holds quarterly meetings to review KPI usage, which it tracks at a detailed level. "If a KPI isn't being accessed, we go back to the owners and see whether they still want it. If not, we remove the KPI," says Martin Summerhayes, program manager.
KPIs are powerful agents of organizational change, but creating effective KPIs is challenging; it is more art than science. While there are many guidelines that can steer teams in the right direction, ultimately, teams must put the KPIs in practice and see what behaviors they drive. Then, the teams need to continually refine and refresh the KPIs to ensure that they are positively affecting organizational change, not undermining it.