How Should We Apply Performance Evaluation to Our Organization?
The moment you copy someone else's system, you fail… Principles of designing performance evaluations tailored to scale, work cycles, and evaluator competency
Why Systems Keep Changing Yet Evaluations Never Do
As the end of the year approaches, a personnel evaluation reform proposal invariably lands on the executive desk. Suggestions emerge to replace relative evaluation with absolute evaluation or to eliminate existing evaluation grades. In some years, OKRs are discussed as a solution, while in other years, ongoing feedback or new performance management systems are proposed as alternatives. Although the names and formats of systems keep changing, the reactions of employees who have experienced the evaluation season often do not change significantly. The questions left for employees who receive their evaluation results tend to focus not on what they should do better moving forward, but on why their evaluation grades and those of their colleagues differed.
This distrust of performance evaluations is difficult to view as a problem unique to specific companies. According to data from Gallup surveying U.S. workers, only 14% strongly agreed that the performance evaluations they receive provide motivation to do better, while 29% strongly agreed that evaluations are fair, and 26% rated them as accurate. Considering that among Fortune 500 Chief Human Resources Officers surveyed by Gallup in 2024, only 2% strongly agreed that their performance management system stimulates employee improvement drive, it is evident that skepticism toward performance evaluations exists not just among evaluated employees, but also among executives and HR responsible for the systems.
The starting point of the problem is likely well ahead of year-end evaluations. In Gallup's survey of U.S. employees, the percentage of respondents who strongly agreed with the item “I know what is expected of me at work” dropped from 61% in 2015 and 56% in 2020 to 47% in 2025, before recovering slightly to 49% in the first half of 2026. Although there are fluctuations over time, this means that even recently, only about half of U.S. workers clearly know what is expected of them. Therefore, before measuring results after the evaluation period ends, it is necessary to check whether the organization sufficiently agreed on what is expected in the first place.
In this regard, explaining the problems of performance evaluation solely through evaluation form items or grading systems is not sufficient. If there is no sufficient agreement on what results members must achieve, how those results connect to organizational goals, and through what procedures goals will be adjusted when the environment changes, it is difficult to gain acceptance for the results no matter how sophisticated an evaluation form is used. Therefore, the first step in reforming the evaluation system is reasonably placed not in choosing a new method, but in checking how clearly expected performance is currently shared within the organization.
Management Insights
When reviewing evaluation system reform proposals, the first question an executive should ask is not whether relative or absolute evaluation is superior. First, it must be verified whether members of our organization can specifically describe the core performance demanded of them and understand how that performance relates to the company's business goals. If it is difficult to clearly answer this question, the process of aligning goals and expectations needs to be supplemented before introducing a new evaluation method or system.
Fairness Must Be Verified Before Evaluation Methods
One piece of data showing what is required for performance evaluations to function as actual performance management tools is McKinsey's 2017 global survey. In July 2017, McKinsey surveyed 1,761 executives across various regions, industries, and company sizes worldwide. Two-thirds of responding companies reported making at least one major change to their performance management systems during the 18 months prior to the survey. Nevertheless, only 38% of respondents evaluated that performance management had a positive impact on both individual employee performance and overall organizational performance. Thus, there was a significant gap between frequently changing systems and the practical effectiveness of performance management.
A particularly noticeable variable in McKinsey's analysis was the perception of system fairness. Among respondents who perceived that the performance management system operated fairly, 60% evaluated the system as effective, whereas among the group that did not see it as fair, the proportion giving the same answer was a mere 7%. Fairness here does not mean that all members should receive the same results or compensation, but is closer to procedural fairness concerning whether evaluation goals and criteria are established rationally, operated under consistent principles, and whether members can receive explanations regarding the process.
McKinsey presented three operational elements connected to this perception of fairness: transparently connecting individual performance goals with business priorities, enhancing managers' capabilities to perform day-to-day coaching, and appropriately reflecting performance differences in compensation. Organizations reporting that they operated all three well showed 84% evaluating their performance management system as effective, whereas in organizations implementing none of the three properly, that proportion stood at a mere 7%. Groups possessing all three elements were approximately 12 times more likely to respond that they had an effective performance management system compared to groups that did not.
Among the three elements, the role of managers appeared particularly important. In McKinsey's survey, 74% of the group answering that managers effectively coached and developed employees evaluated the performance management system as effective, compared to 15% in the group viewing managers as ineffective coaches. However, this data is not an experiment proving causality that a specific system directly generates performance improvement, but a survey showing the correlation between organizational operational methods and perceptions of performance management effectiveness. Therefore, rather than applying the numbers directly to all organizations, it is appropriate to interpret them as evidence showing which operational elements should be checked with priority.
Management Insights
When discussing performance evaluation reform, discussing whether to reduce evaluation grades from five tiers to three or to transition from relative to absolute evaluation first can lead to excessive focus on the external appearance of the system. Instead, it is more useful for diagnosing the practical effectiveness of the system to first check whether members understand the connection between their goals and company priorities, whether managers provided sufficient feedback and coaching during the evaluation period, and how performance differences are reflected in compensation and explanation processes.
The Reality of Korean Organizations: How to View Forced Distribution
To apply overseas survey results directly to Korean companies, the characteristics of domestic evaluation systems must also be examined together. Based on the results of a survey conducted in 2005 [Note: Korean text says 2015] by the Korea Labor Institute and published in 2016 titled "Current Status of Personnel Evaluation Systems," the operational status of evaluations across 501 business entities centered on workplaces with 300 or more employees was analyzed. The study revealed that 67.1% of surveyed entities utilized a forced distribution method allocating headcount per evaluation grade according to pre-determined ratios. Although this is a survey from about a decade ago and cannot be viewed as the latest statistics representing the evaluation methods of all Korean companies recently, it is meaningful as data for understanding how the relative evaluation practices of domestic companies have been operated.
There are operational reasons why the forced distribution method has been used for a long time. It can partially control the tendency of evaluators to assign generally high scores (leniency bias) or the phenomenon of concentrating most members in middle scores. For companies that must distribute limited bonus financial resources by grade, it is also relatively easy to manage the overall scale of compensation. It can also be utilized as a device to prevent result distributions from spreading excessively when the experience and judgment criteria of evaluators differ greatly among departments. However, it is necessary to distinguish that creating a constant distribution and securing the accuracy of individual evaluations are different issues.
From the perspective of members, different problems can arise. If grades are assigned according to pre-determined ratios, an individual's grade can change depending on the performance of other members within the same evaluation group, even if the individual achieved pre-agreed goals. Consequently, a perception can arise that evaluation results are not determined solely by one's own actions and performance. In particular, in jobs where multiple members must jointly create performance, evaluations emphasizing individual relative rankings have a high potential to conflict with collaboration incentives. Therefore, if forced distribution is used, the extent and organizational units to which it applies should be reviewed together.
Conversely, declaring absolute evaluation does not automatically resolve these problems. In situations where an organization has not established specific expected levels in advance or where evaluators interpret the same criteria differently, evaluation inflation—where high scores are given to most employees—can appear. Later, in the process of distributing limited compensation resources, informal ranking adjustments may occur again. Therefore, for organizations intending to adopt absolute evaluation, it is practically important to design evaluator training, evaluation evidence documentation, and calibration procedures to adjust differences in judgment across departments alongside the task of operationalizing evaluation criteria.
Management Insights
It is dangerous to view either relative or absolute evaluation as a universal correct answer applicable to all companies. An organization where evaluators stably apply specific criteria and can pre-agree on expected performance with members can broaden the scope of absolute evaluation. However, if this foundation is insufficient, rather than nominally changing the system to absolute evaluation, an approach that first addresses the problems of the existing evaluation method while clarifying the basis for lower grades and adjustment procedures can be more realistic.
Read Your Organization's Conditions Before Choosing an Evaluation System
The reason why importing success stories from other companies directly is dangerous is that only the form of the system is moved, while the conditions under which it operated cannot be brought along. When designing performance evaluations, it is more rational to first examine conditions such as organization size, work performance cycles, evaluator competency, organizational culture, and business stage, and then determine the evaluation cycle, criteria, and degree of differentiation upon those foundations.
First, the smaller the organization size, the more cautious one needs to be in mechanically applying predetermined grade ratios. For example, applying a ratio of 10% S grade and 20% A grade to a team of five people means one person's grade cannot be divided into an integer, so results will vary depending on rounding or adjustment methods. In such situations, designing how much each member achieved pre-agreed goals and how to connect overall organizational performance to compensation is simpler and easier to explain than creating detailed grade distributions. Conversely, in organizations with hundreds of people, as the number of evaluators increases, the likelihood of departments interpreting evaluation criteria differently grows, increasing the necessity for fixed guidelines and calibration procedures.
Work performance cycles also directly affect evaluation design. Evaluating sales tasks—where monthly sales and contract results are confirmed relatively quickly—and R&D tasks—where results may appear years later—using the same period and method makes it difficult to properly reflect the actual value creation process of each job. Tasks where results appear with a certain time lag, such as content, service planning, and new businesses, require designing how to check agreed execution processes, quality, and interim outputs together rather than evaluating short-term results alone. If goals themselves changed due to environmental shifts, procedures to adjust them prior to evaluation are also necessary.
Gallup has also emphasized periodic progress reviews over one-time year-end evaluations, recommending in past research that progress reviews re-examining performance, goals, priorities, and collaboration relationships be conducted at least every 6 months. As 2024 Gallup data showed that 56% of employees reported the frequency of officially reviewing performance goals with managers as once a year or less, operating interim check-ins tailored to the speed of work changes rather than keeping the evaluation cycle to just once at year-end helps reduce the gap between goals and actual work.
The third condition is evaluator competency. In Gallup's 2024 survey, employees answered most commonly that their performance evaluations are based on manager observation, with that proportion reaching 67%. Ultimately, if managers lack the ability to interpret evaluation criteria or the habit of routinely observing and recording work, year-end evaluations are likely to be influenced by recent memories or events leaving strong impressions. Therefore, prior to advancing the evaluation system, it is necessary to settle practices where managers regularly converse with members about performance and accumulate necessary evidence.
Finally, organizational culture and business stages cannot be separated from evaluation systems. The required level of differentiation can differ between jobs where individual performance is clearly distinguished and jobs where multiple roles jointly create results. Furthermore, growth-stage organizations—where business models are explored rapidly and roles change frequently—and organizations with relatively stable jobs and processes have no reason to use the same goal-setting methods. Therefore, before deciding which evaluation system to introduce, it is necessary to define what behaviors the organization needs to encourage at its current business stage.
Management Insights
What must be checked when benchmarking evaluation systems is not whether another company uses relative evaluation or operates OKRs, but under what conditions that system operates. Because the same evaluation system can produce completely different effects if organization size, work cycles, manager competencies, collaboration methods, and business stages differ, an approach of diagnosing current organizational conditions and selecting and modifying only necessary elements rather than importing systems is realistic.
Performance and Competency Are Not the Same Question
Many organizations evaluate performance and competency separately and then apply certain weights to create a single final score. Combining them as 60% performance and 40% competency is simple from an administrative standpoint, but important differences can disappear in the process of compressing disparate information into a single number. A person with a high level of goal achievement who has not yet sufficiently demonstrated competencies required for a new role, and a person with high professional competencies who failed to create expected results due to external environments or job conditions, might receive the same score, yet the personnel actions required for the two individuals are not identical.
As a practical design proposal to reduce such problems, this article suggests viewing performance and competency on different axes. Because performance is an area confirming to what extent pre-agreed results were generated, it can be relatively directly connected to bonuses or short-term compensation. Competency, on the other hand, is information diagnosing what capabilities are possessed and how much those capabilities are being exerted in current work, making it easier to utilize in connection with development, lateral placement, role expansion, and promotion reviews.
The perspective of “competency utilization rate” proposed by Korea Business Review also originates from this distinction. Simply checking what competencies members possess is not enough; an approach of examining together how much of those possessed competencies are being used in actual work is required. If the level exerted in work is lower than possessed capabilities, rather than immediately judging this as individual competency deficiency, one must diagnose together whether the job and strengths are misaligned, whether decision-making authority is excessively restricted, and whether the current work structure provides opportunities to use those competencies.
Separating performance and competency does not mean ignoring the relationship between the two pieces of information. Rather, maintaining the two axes separately allows for additional examination of why performance is low despite high competency, or whether currently high performance stems from a specific individual's competency or was heavily influenced by market environments or temporary conditions. These differences can be utilized for judgments not only in compensation, but also in talent placement and succession planning.
Management Insights
While the method of combining performance and competency into a single final score has the advantage of simple administration, it can excessively compress information executives need to know about people. Therefore, interpreting performance primarily from the perspectives of results and compensation, and competency from development and placement, and then reviewing the two pieces of information together when necessary can help grasp the talent issues of the organization more concretely.
Reforming Evaluation Systems: What to Do in the First 90 Days
If an organization that has decided to change performance evaluations changes grades, compensation ratios, evaluation forms, and IT systems all at once from the beginning, the system becomes complex while making it difficult to judge why results changed. Therefore, the 90-day implementation plan below is appropriately understood not as a guaranteed standard model from specific research, but as a minimum execution plan proposed by KBR to apply the principles examined earlier—goal clarity, regular performance dialogues, and accumulation of evaluation evidence—to actual organizations.
During the first 30 days, focus is placed on having all members agree with managers on two or three core goals and briefly documenting how each goal connects to business priorities. At this stage, rather than hastily changing existing evaluation grades or compensation methods, it must be verified whether employees and managers share the same picture of what will be judged as performance when the evaluation period ends. If goals themselves are overly abstract, they need to be concretized into measurable results or verifiable outputs.
During the next 30 days, progress check-ins between managers and members are incorporated into actual operational processes. Brief interviews of about 15 minutes once a month can be applied. Here, a cycle of once a month is not an absolute standard necessary for all organizations, but a practical proposal applicable in organizations with fast work changes. The core lies in continuously checking—rather than the interview frequency itself—whether existing goals remain valid, what factors hinder progress, and what support managers need. Having such interim check-ins rather than checking results once at year-end secures opportunities to modify goals when they diverge from reality.
During the final 30 days, actual evaluation and compensation principles are reviewed based on the goals and interview records accumulated during the initial period. Discussing actual cases regarding to what extent goal achievement will be reflected in evaluations, within what scope absolute evaluation and relative adjustments will be used, and how judgment differences among evaluators will be adjusted allows organizations to break away from abstract debates surrounding system names and supplement designs centered on problems occurring in the current organization.
In this process, it is also important not to excessively increase forms and systems. In the early stages, operating only minimal records necessary for goal agreement and progress check-ins while making managers actually conduct dialogues makes it easier to verify effectiveness. Introducing new performance management software before dialogue practices have taken root is likely to only increase fields that must be entered. Above all, when executives do not remain outside the system but show members what their own goals and priorities are, they can signal that the performance management system is not a device merely to control lower-level employees, but an operational method for the entire organization.
Management Insights
The purpose of performance evaluation reform is not to create a new evaluation form, but to build an operating system that clearly agrees on expected outcomes, checks progress, and explains final results based on consistent criteria. If, after confirming that basic processes such as goal agreement, regular check-ins, and accumulation of evaluation evidence actually operate during the first 90 days, grades, compensation, and systems are designed step-by-step, the likelihood of building a performance management system suited to our organization's conditions can be increased compared to simply replicating another company's system.
