OKR guide

    OKRs and performance reviews: how to keep them apart without losing the link

    Keep OKRs and performance reviews connected by evidence, not by a score. Tying ratings or pay to OKR attainment teaches people to set safe targets and hide bad news, which destroys the honesty OKRs depend on.

    By Reuben Chircop, OKR specialist and leadership coach

    Why linking ratings to attainment backfires

    When a rating depends on the percentage reached, people do the rational thing: they set targets they know they can hit. Sandbagging replaces ambition, and the OKRs stop describing what the team is really trying to change.

    Bad news gets hidden until the end of the quarter. A key result drifting off track becomes something to protect rather than something to raise, so the help that could have arrived in week four arrives too late, if at all.

    A stretch target gets treated as a missed promise. OKRs work partly because some key results are deliberately ambitious, and reaching most of the way can be a good result. A rating system reads the same outcome as failure.

    People also start avoiding shared key results they cannot fully control. Cross team work is often where the most important outcomes sit, and it is exactly the work people back away from when their rating depends on someone else's delivery.

    What OKRs can legitimately feed into a review

    Separating them does not mean the OKR work disappears at review time. It is one of the best sources of evidence a manager has, as long as the evidence is about the person's contribution rather than a number.

    • Evidence of contribution. What the person actually did to move the outcomes, and what changed as a result.
    • What they learned. Which assumptions turned out to be wrong, and how that changed their approach.
    • How they handled a key result that went off track. Whether they raised it early, proposed options and adjusted the plan.
    • Collaboration. How they worked with other teams on shared outcomes and dependencies.
    • Quality of judgement and of check-ins. Whether their updates were honest, timely and useful to others.

    What to keep out

    • Percentage attainment as a rating input, in any form or weighting.
    • Ranking or comparing people by key result scores.
    • A key result score used as a bonus formula.

    Each of these turns a planning and learning tool into a measure of the person, and the targets change accordingly.

    Attainment based vs evidence based reviews

    AspectAttainment based reviewsEvidence based reviews
    How targets get setLow enough to be sure of hitting themAmbitious enough to learn from, agreed openly
    Honesty in check-insBad news is held back until it cannot be hiddenRisks are raised early because raising them is safe
    The manager conversationArguing about the numberDiscussing contribution, judgement and learning
    Fairness when the market changesPeople are penalised for events outside their controlJudgement accounts for what changed and how people responded
    Main riskOKRs become a negotiation over safe targetsReviews become vague without a contribution record

    How to run it in practice

    1. Separate the OKR and review calendars. Run quarterly OKR reviews on their own rhythm and hold performance reviews at a different time. When the two meetings coincide, people treat the OKR review as a rehearsal for their rating.
    2. Keep a simple contribution log during the quarter. Each person notes what they did, what changed and what they learned. A few lines after each check-in is enough, and it gives the review real evidence instead of memory.
    3. Use three review questions. What did you contribute? What did you learn? What would you do differently? These questions draw on the OKR work without turning it into a score.
    4. Calibrate on behaviour and impact, not percentages. When managers compare notes, talk about the quality of decisions, collaboration and results achieved. Leave key result percentages out of the calibration discussion entirely.
    5. Tell everyone the rule at the start of the cycle. Explain that OKR attainment will not be used as a rating input before anyone writes a key result. Announcing it at review time is too late, because the targets were already set with the old fear in mind.

    Signs attainment is leaking into ratings

    These signs are prompts for a conversation, not proof. Ask people what they expect the OKRs to be used for before deciding on a fix.

    SignLikely causeFix
    Every key result lands at exactly 100%Targets are being set to be safeRestate the rule and ask for one genuine stretch key result per objective
    People refuse shared key resultsThey fear being judged on outcomes they cannot fully controlName one owner for shared outcomes and review contribution, not the shared score
    Check-ins turn green the week before reviewsPeople expect the status to be read as a ratingMove the review date away from OKR check-ins and ask managers not to reference status colours
    Managers quote OKR percentages in review formsThe review template invites itReplace the field with the three review questions and remove percentage fields

    FAQ

    Not through the percentage achieved. A bonus formula built on key result scores rewards safe targets and punishes honest stretch. If bonuses exist, base them on a broader judgement of contribution and impact, where evidence gathered during the OKR cycle can inform the conversation without deciding the number.

    A sales quota is a compensation target, and that is fine as such. It is agreed, measured and paid against in its own system. The problem comes when a quota is dressed up as an OKR. Keep the quota in the compensation plan, and use OKRs for the changes the sales team is trying to make, such as entering a new segment or shortening the sales cycle.

    Yes. Progress and confidence are useful context for regular conversations about priorities, blockers and support. They should never become a score attached to a person. The manager's question is what is happening and what would help, not what percentage someone reached.

    Offer a practical compromise: let the review draw on evidence of contribution from the OKR cycle, such as what someone did, what changed and what they learned, but never on the attainment percentage. Agree that rule in writing and announce it at the start of the cycle, so nobody adjusts their targets in anticipation.

    Practice varies, but many organisations that use OKRs deliberately keep them separate from compensation, for the reasons on this page. Whatever others do, the test for your own organisation is simple: are people still setting ambitious targets and reporting bad news early? If not, the link is probably too tight.

    If writing the OKRs themselves is the slow part, the free OKR generator from OKRgen drafts one objective with three key results as a starting point.