People Analytics

When organizations begin writing OKRs, they first tend to cling to the wording. They try to make the Objective sound more impressive and turn the Key Results into more precise numbers. But good OKRs are not a matter of writing skill. They are a matter of deciding what to give up and what to focus on.

As discussed in the previous articles, KPIs are indicators for observing the state of an organization, while OKRs are a device for asking what direction the organization intends to change during this period. If this difference is not reflected in the design of Objectives and Key Results, OKRs quickly become a task list or an evaluation sheet. Good OKRs ask not “what will we do?” but first “what change must actually happen during this period?”

An Objective is not an impressive sentence, but a device for deciding what to give up

Google’s OKR playbook describes Objectives as “Whats.” This means they should show what is to be achieved and what intent and direction are involved. It also says that a successfully achieved Objective must provide clear value to the organization. In this sense, an Objective is not a slogan but a sentence of choice.

For example, the phrase “create the best employee experience” sounds good, but it does not show a choice. It does not reveal which employee experience matters, why it matters now, what will be given up, or where attention will be focused. By contrast, “redesign the onboarding experience to reduce first-90-day attrition risk for new hires” is narrower. That is why it is more operational.

A good Objective turns broad desire into a narrow priority. If an HR department writes that it will change recruiting, training, evaluation, organizational culture, labor relations, and HR Tech all in one quarter, it is the same as choosing nothing. An Objective must set the direction in which the organization will move most significantly during this period and draw a line that leaves other work as KPIs or ordinary operations.

The difference becomes clearer in practice. If an HRD team writes “improve satisfaction with leadership training” as its Objective, it is likely to remain at the level of improving training operations. By contrast, if it writes “reduce early 90-day management failures among newly appointed team leaders,” it will look together at onboarding, 1:1 meetings, feedback quality, and signals of team member attrition. A good Objective is not a pretty sentence; it reveals the scene the organization actually needs to change.

Key Results should be visible results of change, not an execution list

Google’s playbook describes Key Results as “Hows,” but it also emphasizes that KRs should describe results, not activities. The original text warns that KRs containing words such as consult, help, analyze, and participate may be signals that they are describing activities. This applies directly to HR practice.

“Conduct three manager training sessions” may look like a KR, but in reality it is an activity. “Carry out interviews with all employees” is the same. Training and interviews may be important. However, unless they show what change they created, they are unlikely to become true Key Results for OKRs. Whether managers’ feedback quality improved after training, whether the retention risk of key talent fell after interviews, or whether the time for new hires to reach productivity decreased is closer to a result.

What Matters explains that an OKR is usually composed of 3 to 5 Key Results under one Objective. This number is not merely a format but a quality standard. If there are too many KRs, they become a task list rather than results. Items that cannot fit within 3 to 5 may not be the core change of this Objective.

HR department OKRs should ask about organizational behavior change rather than activity volume

OKRs in HR departments are especially prone to drifting toward activity volume. Recruiting teams can easily write the number of job postings and interviews, HRD teams the number of training sessions and completion rates, and organizational culture teams the number of campaigns and participation rates. These indicators may be useful as KPIs. But to become KRs in OKRs, they must show behavior or results that changed after the activity.

For example, if the recruiting team’s Objective is “increase decision-making speed for hiring in critical roles,” a KR could be written as “reduce the median time from interview completion to offer decision from 7 days to 4 days.” If the HRD team’s Objective is “reduce early management failures among new leaders,” a KR could be designed as “increase the rate of 1:1 feedback within 60 days of new leader assignment from 40% to 85%.”

The same applies to organizational culture teams. “Run five campaigns” is less like an OKR than “increase the share of issues converted into action items in team retrospective meetings from 30% to 60%.” What matters is not evidence that many good activities were carried out, but evidence that the organization’s behavior actually changed.

A good OKR meeting is not a meeting that adds goals, but one that reduces them

Atlassian’s OKR guide presents a flow for defining 1 to 3 Objectives and setting 3 to 5 Key Results for each Objective. Although it has the limitation of being vendor material, these numbers are practically useful. They show that an OKR meeting should not be a place that keeps adding goals, but a place that reduces them.

Goal-setting meetings in Korean companies often flow in a way that reflects the demands of every department. Executive priorities, division head instructions, requests from the field, and existing KPIs enter one document together. When this happens, OKRs become the outcome of stakeholder coordination rather than a tool for creating strategic focus.

A good OKR meeting requires three questions. First, what must be changed during this period? Second, what will be given up for this Objective or managed only as a KPI? Third, do the KRs show results rather than activities? If a goal cannot pass these questions, it should be reduced rather than written in further detail.

The failure factors in the next installment appear in operating rhythm rather than writing methods

Creating good Objectives and Key Results does not mean OKRs will work. Even good sentences become end-of-quarter evaluation materials if there is no operating rhythm. If check-ins turn into reporting meetings, if there are no criteria for changing goals, and if leaders do not remove obstacles, OKRs become extra work for the field.

Therefore, the final step in writing OKRs is not reviewing the wording, but agreeing on operations. Who will check progress, and at what cadence? Who has the authority to revise goals when KRs become unstable? Who will coordinate conflicts between departments around shared goals? Will achievement rates be viewed as evaluation scores or as material for performance conversations?

OKRs are not a technique for writing good sentences, but a way to reveal the organization’s choices. Objectives show what has been chosen, and Key Results confirm whether that choice has turned into actual results. If this standard cannot be maintained, OKRs become another name for KPIs. If it can be maintained, OKRs can move performance management conversations from activity volume to evidence of change.

The confusion that most often arises in companies adopting OKRs concerns their relationship with KPIs. The name changes to OKR, but in many cases the actual operation is no different from a KPI table. Existing metrics such as revenue, cost, time to hire, training completion rate, and turnover rate are copied as they are, and the quarter-end achievement rate is interpreted as evaluation material. In this approach, OKR becomes not the language of performance management transformation but new packaging for the existing evaluation sheet.

The difference between OKRs and KPIs is not “which one is newer.” They are different questions about organizational performance. KPIs ask whether current operations are healthy. OKRs ask what the organization intends to change during this period. When this distinction becomes blurred, companies turn every number into a goal and every goal into an evaluation score, ultimately leaving employees with only more reporting burden.

KPIs show the state of the organization, while OKRs ask what direction it wants to change

KPI.org describes a KPI as a key performance indicator, a metric used to evaluate performance or success. What matters in this definition is that a KPI is a device for observing the operating condition of an organization. For example, a recruiting team’s average time to hire, a sales team’s conversion rate, a customer support team’s response time, and an HRD team’s training completion rate are closer to health indicators that the organization must manage steadily.

What Matters explains the difference between OKRs and KPIs more directly. OKRs are metrics that measure change, while KPIs are metrics that measure health. This distinction is very important in performance management practice. Health indicators must be monitored continuously. But not every health indicator needs to become an OKR. OKR is a way to select, from among them, the priorities the organization actually wants to change in the current quarter.

For example, turnover rate can be a KPI because it is needed to check organizational stability every month. But a KR such as “reduce the first-year attrition rate in critical job groups from 12% to 8%” moves closer to an OKR. It contains not a simple observation metric but a result the organization intends to change within a specific period. The same number is managed differently depending on where it is placed.

The same number carries different responsibility when placed in a KPI versus an OKR

What Matters explains that an OKR is usually composed of 3 to 5 Key Results under one Objective. This number is less about format than about the meaning of selection. An organization may manage dozens of indicators, but the Key Results included in an OKR must be limited. Without limits, there is no priority.

A number placed in a KPI mainly creates “responsibility to maintain the state or detect deterioration.” For example, a 95% training completion rate is useful for seeing whether training operations are proceeding as planned. But that figure alone does not show whether employees’ capabilities have actually changed in their work. By contrast, an OKR KR asks, “What change actually occurred?” The more important questions become whether the time for new sales representatives to reach their first contract after training has decreased, whether managers’ feedback quality scores have improved, or whether the success rate of internal mobility applicants in role transitions has increased.

Asana’s explanation comparing OKRs and KPIs presents a similar axis. It explains that OKRs combine ambitious goals with measurable results to drive change, while KPIs monitor continuous performance and operational health. Although it has the limitation of being vendor material, the practical distinction itself is useful in the performance management field. Based on this distinction, HR should first separate “Is this number an indicator we need to keep watching, or is it a result we need to change this quarter?”

When Google-style OKRs enter Korean companies unchanged, they can become evaluation sheets

The Google OKR playbook describes OKR as a process for communicating, measuring, and achieving ambitious goals. It also distinguishes the nature of goals, such as committed OKRs, aspirational OKRs, and cross-team OKRs. What Korean companies should take from this is not the template but the method of distinction.

A committed OKR is close to a goal whose achievement has been promised. For such goals, execution responsibility is relatively clear. An aspirational OKR is more challenging. Its chance of success is not completely guaranteed, but the organization sets it to test new possibilities. A cross-team OKR involves contributions from multiple departments. Problems arise when these three types are evaluated in the same way.

In Korean companies, the moment OKRs turn into evaluation sheets usually begins here. When every OKR is converted into an individual score and achievement rates are directly linked to rewards, employees choose safe goals rather than ambitious ones. Shared goals across departments can turn into disputes over responsibility. Goals from which the organization could learn through failure remain as records of non-achievement. To use OKRs as a tool for transforming performance management, the work of separating responsibility and interpretation by goal type must come first.

What HR must decide first is not metric names but operating rules

In organizations that use OKRs and KPIs together, what HR must decide is not the label. More important are the operating rules. Which indicators will remain on the KPI dashboard? Which indicators will be elevated to KRs in OKRs? Which goals will be viewed as evaluation reference material, and which goals will be viewed as evidence for learning and strategic adjustment? Without answers to these questions, OKRs are immediately absorbed into the existing evaluation system.

The first rule is to distinguish the purpose of each indicator. Indicators that look at stable operations are better left as KPIs. Strategic changes to be made during the current period should be handled as OKRs. The second rule is to record the difficulty of goals. If committed goals and aspirational goals are not distinguished, achievement rates become numbers that are difficult to interpret. The third rule is to define the nature of check-in meetings. OKR check-ins should not be reporting meetings but meetings for adjusting priorities and removing obstacles.

The fourth rule is to decide the distance from evaluation. It is difficult to exclude OKR results completely from evaluation, but directly converting achievement rates into scores is also risky. Especially for aspirational OKRs and shared OKRs, the judgment process, learning, collaboration responsibility, and the leader’s coordination role should be considered together rather than only the achievement rate.

The question for the next installment is not how to write good OKRs, but what to give up

Once the difference between OKRs and KPIs has been distinguished, the next question moves to how to write them. But writing good OKRs is not merely a matter of sentence technique. A good Objective is not a stylish phrase; it reveals a choice. The organization must decide what it will not do this quarter, which numbers it will only observe, and which results it must change.

A good Key Result is not a list of activities either. “Conduct training,” “hold interviews,” and “run meetings” may be action items, but they are not results. An OKR KR must show what actually changed for the organization or its members after the activity. If training was conducted, it must be confirmed what changed among capability, behavior, mobility, and performance. If interviews were held, results such as retention rate, execution rate of growth plans, and manager feedback quality should follow.

Ultimately, the difference between OKRs and KPIs is a matter of changing the language of performance management. KPIs show the state that the organization must continue to observe. OKRs reveal the direction the organization has committed to changing now. If the two tools cannot be distinguished, performance management becomes more complicated. Conversely, if they can be distinguished, HR can move from being the department that manages evaluation sheets to the role of designing the rhythm of strategy execution.

Whenever more companies bring OKRs back into discussion, a familiar scene repeats itself. A new template is created, departmental goals are entered, and quarterly review meetings are scheduled. But after a few months, OKRs either become another name for the existing KPI table or are pushed aside as reference material to pull out during evaluation season.

The issue is not whether OKRs themselves are in fashion. In the 2026 performance management discussion, a more important question lies elsewhere. It is not how well a company writes its goals, but whether it can explain strategic priorities, execution accountability, feedback rhythm, and the relationship between evaluation and rewards within a single system. At this point, OKRs move from being a goal management template to a question of the performance management operating system.

The reason OKRs matter again lies not in templates but in the problem of alignment

What Matters explains OKRs as Objectives and Key Results, that is, the combination of objectives and key results. The important point here is not “writing goals,” but the explanation that OKRs “track progress, create alignment, and increase engagement around measurable goals.” OKRs are closer to a mechanism for checking whether the organization is looking in the same direction than to the goal statement itself.

If this difference is missed, OKRs quickly become another name for KPIs. KPIs are indicators used to continuously monitor operational performance. Numbers that must be managed consistently, such as revenue, time to hire, turnover rate, training completion rate, and customer response time, belong here. By contrast, OKRs are a commitment for judging what the organization wants to change over a given period and whether that change has actually occurred. Even when the same numbers are used, their purpose is different.

What Matters explains that an OKR is usually composed of three to five Key Results under one Objective. It also states that OKRs are reviewed over a set period, often quarterly. The message this structure conveys is clear. OKRs are not a container for every task, but an editing device that leaves only the changes the organization truly needs to focus on during that period.

What the Google case shows is a shared direction more than high goals

Google, often cited as a representative OKR case, describes OKRs in its internal OKR playbook as “a process for communicating, measuring, and achieving high goals.” This sentence shows that OKRs are not simply a goal entry template but a way of communicating. More important than the fact that goals are high is whether multiple teams move around the same priorities.

The Google playbook does not treat OKRs as a single type. It distinguishes committed OKRs, aspirational OKRs, and cross-team OKRs. Committed OKRs are close to goals that have literally been promised for achievement, while aspirational OKRs are closer to challenging goals whose path has not been fully confirmed. Cross-team OKRs deal with goals to which multiple organizations must contribute together.

This distinction offers important implications for Korean companies as well. If all OKRs are viewed through the same evaluation standard, challenging goals disappear. Conversely, if all OKRs are left loose under the name of “challenge,” execution accountability becomes blurred. Therefore, what must be decided first in OKR operations is not the wording of the goal but the nature of the goal. Whether it is a commitment that must be achieved, a challenge for learning, or a problem that several departments must solve together should change the way it is reviewed.

However, copying the Google case as it is would be risky. The playbook itself assumes that it is Google’s approach and that other organizations’ approaches may differ. Companies with different scale, industries, organizational cultures, evaluation systems, and leadership styles will not get the same effect simply by bringing in the same template. What should be taken from the case is not the document format, but the way goals were made into a language of conversation and alignment across organizations.

The Microsoft case reveals that OKRs are an operating rhythm more than software

Microsoft Learn’s Viva Goals document explains that the OKR framework is used to connect teams to the organization’s strategic priorities, timelines, and goals. It also states that business outcomes are driven by regularly checking in on OKR status and progress. This explanation shows how OKRs are productized from an HR Tech perspective.

But here, too, the core issue is not the tool. Software can make goals visible, record progress, and support the check-in rhythm. However, it cannot decide on behalf of the organization which goals are strategic priorities, who will coordinate when goal conflicts arise, or how achievement rates should be interpreted in evaluation and rewards.

The message the Microsoft case gives HR practitioners is clear. Operational questions must be organized before introducing an OKR tool. First, what are the rules for connecting company-wide goals and team goals? Second, who approves goal changes during the quarter? Third, are check-in meetings reporting meetings or obstacle-removal meetings? Fourth, is the OKR achievement rate an evaluation score or evidence for a performance conversation?

Without these questions, implementing a system merely turns OKRs into a faster input form. Conversely, in an organization where the questions have been clarified, even a simple spreadsheet can become an operating system. The success or failure of OKRs is closer to decision-making rules and meeting rhythms than to purchasing a solution.

For Korean companies, the issue is the allocation of accountability more than whether OKRs are reflected in evaluation

The point at which OKR discussions become difficult in Korean companies is evaluation and rewards. Practitioners often ask, “Should OKRs be reflected in evaluation?” This question matters, but if the discussion immediately moves to pros and cons, it becomes too narrow. What should be asked first is, “What kind of accountability is this OKR meant to explain?”

Committed OKRs are close to promised execution accountability. In this case, whether they were achieved and why they fell short can become important material in performance conversations. Aspirational OKRs have a strong character of uncertain challenge and learning. If only the achievement rate is evaluated, employees will choose safe goals. Cross-team OKRs involve high interdependence among departments. If such goals are simply distributed into individual evaluations, avoidance of responsibility may increase more than collaboration.

Therefore, the relationship between OKRs and evaluation and rewards is difficult to fix as a single principle. What Korean companies need to decide is not “reflect them” or “do not reflect them,” but interpretation rules by OKR type. Some OKRs become evidence of performance accountability, some become evidence for learning and strategic adjustment, and some become material for checking accountability for coordination across organizations.

The role of the HR department also changes here. HR should become an operator that designs goal types, check-in rhythms, leaders’ feedback standards, ways of using OKRs as evaluation references, and exception-handling principles, rather than an administrator that distributes OKR forms. If the performance management system is not to operate only during evaluation season, OKRs must be connected to a conversation structure that repeats throughout the quarter.

This series tracks OKRs as a problem of connecting systems, leadership, and data

HR Lens does not treat this OKR series as a simple explanation of concepts. The first axis is the difference between KPIs and OKRs. If KPIs are stable operational indicators, OKRs deal with strategic change and priorities. The two are not substitutes for each other, but a relationship that must be designed together.

The second axis is how to write them. A good Objective is not a stylish sentence but a statement that reveals a choice. A good Key Result shows an outcome, not an activity. “Conduct training” or “operate meetings” is closer to an execution list. It becomes a KR when it shows what behavior, capability, performance, or movement changed after the training.

The third axis is failure factors. Organizations where OKRs fail usually create too many goals, see management priorities change frequently, and turn check-in meetings into reporting meetings. Many also leave only achievement rates behind without clarifying the relationship with evaluation and rewards.

The fourth axis is leadership and data. OKRs are an HR system and, at the same time, a leadership system. If leaders do not clarify priorities and do not coordinate goal conflicts across departments, OKRs remain a burden on the field. In the future, People Analytics and AI may summarize goal progress and detect risk signals, but responsibility for interpretation will still remain with people and organizations.

The standard for this series is one thing. It is not whether OKRs were introduced, but whether OKRs changed the organization’s performance conversations. It is not whether more goals were entered, but whether it became clearer what to give up and what to focus on. If this question cannot be answered, OKR becomes just another system name. If it can be answered, OKRs can become an opportunity to move the center of performance management from evaluation to execution and learning.