Performance management: Why keeping score is so important, and so hard

Performance management: Why keeping score is so important, and so hard

By Raffaele Carpi, John Douglas, and Frédéric Gascon

The elements of a good performance-management system are simple, but integrating them into a business’s fundamental operating system is more difficult than it seems.

Effective performance management is essential to businesses. Through both formal and informal processes, it helps them align their employees, resources, and systems to meet their strategic objectives. It works as a dashboard too, providing an early warning of potential problems and allowing managers to know when they must make adjustments to keep a business on track.

Organizations that get performance management right become formidable competitive machines. Much of GE’s successful transformation under former CEO Jack Welch, for instance, was attributed to his ability to get the company’s 250,000 or so employees “pulling in the same direction”—and pulling to the best of their individual abilities. As Henry Ford said, “Coming together is a beginning; keeping together is progress; working together is success.”

Yet in too many companies, the performance-management system is slow, wobbly, or downright broken. At best, these organizations aren’t operating as efficiently or effectively as they could. At worst, changes in technologies, markets, or competitive environments can leave them unable to respond.

Strong performance management rests on the simple principle that “what gets measured gets done.” In an ideal system, a business creates a cascade of metrics and targets, from its top-level strategic objectives down to the daily activities of its frontline employees. Managers continually monitor those metrics and regularly engage with their teams to discuss progress in meeting the targets. Good performance is rewarded; underperformance triggers action to address the problem.

Where do things go wrong?

In the real world, the details of performancemanagement systems are difficult to get right. Let’s look at a few common pitfalls.

Poor metrics

The metrics that a company chooses must actually promote the performance it wants. Usually, it can achieve this only by incorporating several of them into a balanced scorecard. Problems arise when that doesn’t happen. Some manufacturing plants, for example, still set overall production targets for each shift individually. Since each shift’s incentives are based only on its own performance, not on the performance of all shifts for the entire day, workers have every incentive to decide whether they can complete a full “unit” of work during their shift.

If they think they can, they start and complete a unit. But if they don’t, they may slow down or stop altogether toward the end of the shift because otherwise all of the credit for finishing their uncompleted work would go to the following shift. Each shift therefore starts with little or no work in process, which cuts both productivity and output. A better approach would combine targets for individual teams with the plant’s overall output, so workers benefit from doing what they can to support the next shift as well as their own.

Poor targets

Selecting the right targets is both science and art. If they are too easy, they won’t improve performance. If they are out of reach, staff won’t even try to hit them. The best targets are attainable, but with a healthy element of stretch required.

To set such targets, companies must often overcome cultural barriers. In some Asian organizations, for example, missing targets is considered deeply embarrassing, so managers tend to set them too low. In the United States, by contrast, setting a target lower than one achieved in a previous period is often deemed unacceptable, even if there are valid reasons for the change.

Lack of transparency

Employees have to believe their targets encourage meaningful achievement. Frequently, however, the link between individual effort and company objectives is obscure or gets diluted as metrics and targets cascade through the organization. Different levels of management, in an attempt to boost their own standing or ensure against underperformance elsewhere, may insert buffers into targets. Metrics at one level may have no logical link to those further up the cascade.

In the best performance-management systems, the entire organization operates from a single, verified version of the truth, and all employees understand both the organization’s overall performance and how they contributed to it. At the end of every shift at one company in the automotive sector, all employees pass the daily production board, where they can see their department’s results and the impact on the plant’s performance. The company has linked the top-line financial metrics that shareholders and the board of directors care about to the production metrics that matter on the ground. Frontline employees can see the “thread” that connects their daily performance with the performance of their plant or business unit (Exhibit 1).

Leaders adapt and cascade performance indicators to all staff levels.

A senior leader at another manufacturer aligns the whole organization around a shared vision through quarterly town-hall meetings for more than 5,000 staff. Managers not only share the company’s financial performance and plant-specific results but also introduce new employees, celebrate work anniversaries, and recognize successful teams. Most important, if targets are missed, the senior leader acts as a role model by taking responsibility.

Lack of relevance

The right set of metrics for any part of a business depends on a host of factors, including the size and location of an organization, the scope of its activities, the growth characteristics of its sector, and whether it is a start-up or mature. To accommodate those differences, companies must think both top-down and bottom-up. One option is the hoshin-kanri (or policy-deployment) approach: all employees determine the metrics and targets for their own parts of the organization. Employees who set their own goals tend to have a greater sense of ownership for and commitment to achieving them than do those whose goals are simply imposed from above.

Lack of dialogue

Performance management doesn’t work without frequent, honest, open, and effective communication. Metrics aren’t a passive measure of progress but an active part of an organization’s everyday management. Daily shift huddles, toolbox talks, after-action reviews, and the like all help to engage team members and to maintain a focus on doing what matters most. Applying the “plan–do–check–act” feedback loop, based on pioneering research from Charles Shewhart and W. Edwards Deming, helps teams learn from their mistakes and identify good ideas that can be applied elsewhere. And in many high-performing companies, supervisors act as coaches and mentors. One-on-one sessions for employees demonstrate concern and reinforce good habits at every stage of career development.

Lack of consequences

Performance must have consequences. While the majority of employees will never face the relentless “win or leave” pressure typical of professional sports, weak accountability tells people that just showing up is acceptable.

Rewarding good performance is probably even more important than penalizing bad performance. Most companies have various kinds of formal and informal recognition-and-reward systems, but few do enough of this kind of morale building, either in volume or frequency. In venues from lunchroom celebrations to town-hall announcements, employee-of-the-month and team-achievement awards are invaluable to encourage behavior that improves performance and keeps it high. One COO at an industrial-goods company keeps a standing agenda item in the monthly business review for recognizing the performance of individuals and teams. Employees on the list may find a gift waiting at home to thank them (and their families) for a job well done.

Lack of management engagement

The words of Toyota honorary chairman Fujio Cho—“Go and see, ask why, show respect”—are now famous as basic lean-production principles. Yet in many companies, senior managers rarely visit plants except during periodic business reviews, and they appear on the shop floor only when a major new capital improvement is to be inspected.

Management interactions with frontline personnel are an extremely powerful performance-management tool. They send a message that employees are respected as experts in their part of the business, give managers an opportunity to act as role models, and can be a quick way to solve problems and identify improvements.

One company’s machinery shop, for example, had developed such a reputation for sloppiness and missed deadlines that managers suggested outsourcing much of its work. When a senior manager was persuaded to visit the workshop, he was appalled at the dirty, cluttered, and poorly maintained environment. Employees reported chronic underfunding for replacement parts and tools, and asked the manager what it would take to save their jobs. He told them to “clean up the shop and give me a list of what needs to be fixed.” Both sides lived up to their commitments, and in less than a year the shop became a reference case for efficiency within the company.

Building a strong performance-management system

The best companies build performance-management systems that actively help them avoid these pitfalls. Such systems share a number of characteristics.

Metrics: Emphasizing leading indicators

Too often, companies measure and manage performance through lagging indicators, such as compliance with monthly output or quality targets. By the time the results are known, it is too late to influence the consequences. The best companies track the same metrics—but also integrate their performance-management systems into critical process inputs. Industrial Internet technologies, such as the SCADA1 architecture and distributed-control systems, let manufacturing staff know within minutes (or seconds) about variations in performance, even in remote parts of a plant. That lets people react long before the variation undercuts output or quality.

Some changes require almost no investment in technology. At the end of each workday, for example, production and functional teams can complete a checkout form assessing how it went. A combination of quantitative and qualitative metrics and simple graphics (such as traffic lights and smiley faces) provides an easy, highly effective tool for identifying and correcting issues or problems before the next day’s work begins.

As performance-management systems evolve, the metrics they use will become more complex, incorporating continuous rather than discrete variables: “everyone showed up on time today” will become “the team achieved 93 percent on the schedule-performance index using 90 percent of the labor-performance index.” The extra detail better informs decisions such as whether to add more labor to meet a delivery date or to push out a schedule for delivery.

Sustainability: Standard work and a regular heartbeat

Regardless of changes to metrics and targets, the best companies keep the cadence of meetings and reviews constant, so they become an intrinsic part of the rhythm of everyday operations (Exhibit 2).

A regular performance-review cadence allows issues to be identified and resolved in an appropriate time frame.

The emphasis on regular, standardized processes goes beyond explicit performance-management activities and extends deep into every aspect of a company’s operating models. Standard work, for example, is based on three simple rules. First, there should be a standard for all activities. Second, everyone must have the knowledge and ability to meet that standard. Finally, compliance with it must be monitored and measured.

In many functions, the business cycle forces a regular rhythm or cadence: the weekly payroll, the monthly accounting close, or the quarterly inventory review. Good companies take advantage of these requirements to define a few central metrics, such as cycle times and accuracy, thus driving continuous improvement across every function.

As part of a lean-manufacturing excellence program, one industrial-commodities company encourages employees to indicate “what went well today, what didn’t go well today, what management can do to help” on their productionarea boards every day. Supervisors collect the information on index cards and post them on a lean-idea board. Representatives of each function meet with the plant manager every morning and accept or reject the cards or return them for more information. Every accepted card gets an owner and timeline for completion. Company leaders estimate that the boards generate at least $2 million a year in cost savings or higher output—but the impact on employee morale and engagement is “priceless.”

A checklist or standard operating procedure that defines the steps and sequences for every key process usually enforces standard work. In employee onboarding, for example, one company noted that small details—assigning email addresses, telephone numbers, and software and hardware access—were especially important for retaining employees early in their tenures. A checklist is now at the front of each new hire’s personnel file, with a copy in the supervisor’s file. The performance reviews of supervisors now assess how well they handled the onboarding of new employees, and everyone who resigns completes a mandatory exit interview.

Continuous improvement: Standard work is for leaders too

Standard work is essential at all levels of an organization, including the C-suite and senior management in general. Standard work for leaders forces a routine that, while uncomfortable at first, develops expectations throughout an organization. It is those expectations, along with specific metrics, that ultimately drive predictable, sustainable performance.

One global resources company now requires managers to demonstrate that they spend 50 percent of their time on a combination of coaching their people and attending safety briefings, shift huddles, improvement reviews, and production meetings. To free up time, other meetings are scheduled only on one day a week— and conference rooms no longer have chairs.

Taking this approach even further, every autumn a field-services organization commits itself to a comprehensive, enterprise-wide calendar for the entire following year. The calendar sets dates for all conferences, monthly and quarterly management meetings, formal performance reviews, and succession-plan meetings, as well as training and development opportunities. All agendas are fixed, and all meetings are subject to strict time limits. There is little need for additional leeway because internal reporting follows tight guidelines for transparency and timeliness: financial results are published internally every month, while data on the performance of teams and units in meeting annual incentive-plan goals are updated and published monthly on bulletin boards.


Most industrial companies have access to rich data on the performance of their operations. The technological advances associated with increasing use of automation, advanced analytics, and connected devices mean that this resource constantly improves. But how can organizations best use their data? A crucial part of the answer is instant feedback loops, daily performance dialogues, and routine performance reviews. Maintaining the willingness and ability to hardwire these performance-management processes into the rhythm of daily work isn’t sexy—but over the long run, it’s the most effective route to real, sustainable performance improvements.

About the author(s)

Raffaele Carpi is a partner in McKinsey’s Lisbon office, John Douglas is an alumnus of the Houston office, and Frédéric Gascon is a senior vice president of McKinsey Recovery and Transformation Services and is based in the Montréal office.

Related

The great re-make: Manufacturing for modern times

The great re-make: Manufacturing for modern times

This 21-article compendium gives practical insights for manufacturing leaders looking to keep a step ahead of today’s disruptions.
More on Operations
Survey

How to create an agile organization

Article - McKinsey Quarterly

Where is technology taking the economy?

Commentary - McKinsey Quarterly

Creating an innovation culture

Article - McKinsey Quarterly

Culture for a digital age