How to Run a Retrospective When OKRs Are Missed

TL;DR. When OKRs are not achieved, review the gap between the intended result and the actual outcome before setting the next cycle. Check the measure, revisit the original assumptions, examine execution choices and identify external changes. Use the retrospective to decide what to continue, change, test or stop, with an owner and a review point for each decision. Do not automatically lower the target or treat the missed result as evidence of individual failure.

What should a missed-OKR retrospective decide?

The meeting should answer a practical question: what will we do differently because of what we learned?

Review each missed key result separately before drawing conclusions about the objective. One result may have suffered from a delayed launch; another may have tested an assumption that proved wrong.

Keep the result and the explanation distinct. Record the miss as it happened. Then examine why it happened without rewriting the target or turning every explanation into an excuse.

Leave with decisions about the objective, the approach, the resources and the measurement. “Communicate better” is too vague. “Review the shared dependency with its owner before committing the next launch” gives someone an action.

What should you prepare before the meeting?

Ask the objective owner to assemble a short evidence pack:

• The original objective, key results, baselines and targets.
• The actual result and its measurement source.
• The assumptions behind the target and planned approach.
• A timeline of launches, delays, priority changes and decisions.
• Relevant customer feedback and evidence of external changes.
• Dependencies and capacity commitments, including changes during the cycle.

Use the definitions agreed at the start. If the definition changed, show both versions and explain the difference.

For example, do not compare an initial target for all new accounts with a final result covering only a selected segment. Resolve the comparison before interpreting performance.

If the key result measures completed activity, ask whether it captures the intended outcome. OKRs vs KPIs offers a useful starting point for separating a change you want to achieve from an ongoing measure you monitor.

Which agenda produces decisions?

Use the following sequence as a working agenda. Give each missed result its own discussion, then review dependencies across the objective.

Step	Question to answer	Required output
Confirm the result	What did we commit to, and what happened?	An agreed statement of the gap
Revisit assumptions	What had to be true for the plan to work?	Supported, rejected and unresolved assumptions
Review execution	What did we choose, deliver, delay or stop?	Specific choices and their observed effects
Examine external changes	What changed outside our control?	Evidence of the change and the response
Choose the response	What should continue, change, be tested or stop?	A small set of decisions
Close the review	Who will act, and when will we check?	Owners, review points and evidence to collect

Start with a brief written reflection from each participant before open discussion. Ask people to distinguish observations from interpretations: “The launch moved” is an observation; “the delay caused the miss” needs further evidence.

Record unresolved disagreements. Where the evidence cannot settle a question, define a test instead of forcing agreement.

How do you separate assumptions, execution and external changes?

Use these categories as prompts, not competing explanations. A missed result may involve all three.

Assumptions concern what the plan expected to be true. Did customers have the problem you intended to solve? Did the proposed change address it? Was the target based on evidence or an untested expectation?

Execution choices concern what the team did with the information and capacity available. Was the work delivered early enough to observe a result? Were dependencies agreed? Did competing work displace the objective? When evidence challenged the plan, what changed?

External changes concern conditions the team did not control. A customer may have postponed a purchase, or a required service may have become unavailable. Ask what changed, when it became visible and how the team responded.

Avoid hindsight as the standard of judgment. Ask whether a decision was reasonable given what was known then, and what signal should prompt a different decision next time.

Accountability still matters. Name decisions, responsibilities and missed commitments precisely. Discuss individual performance separately when needed; do not infer it from the objective’s final score alone.

What does a useful worked example look like?

Illustrative example: a SaaS team sets a key result to increase newly activated accounts from 40 to 60 per cycle. It finishes at 46.

The team’s initial explanation is that onboarding shipped late. The retrospective checks that account:

• Assumption: customers would complete setup once the instructions became clearer.
• Execution: the revised onboarding launched near the end of the cycle.
• Evidence: support conversations also identified missing access to customer data as a setup blocker.
• External change: a prospective customer group postponed its rollout.

The late launch limited the opportunity to observe results. However, clearer instructions did not address every identified blocker. The postponed rollout explains part of the gap, but the team cannot yet show how much.

The decisions are therefore specific:

• Continue observing the revised onboarding before claiming an effect.
• Test an assisted setup path for accounts blocked by data access.
• Confirm rollout readiness before using that group in the next forecast.
• Reassess the target using the eligible account base and available delivery time.

The team records the miss unchanged. It also avoids claiming that either the launch delay or customer postponement explains the whole result.

Should you lower the target or carry the OKR forward?

Choose the response from the evidence, rather than from the discomfort of missing.

Response	Use it when	What must change
Keep the objective and target	The outcome still matters and the target remains defensible	The approach or a known execution constraint
Revise the target	Evidence changes the baseline, eligible population or feasible scope	The documented basis for the target
Change the key result	The measure does not adequately represent the intended outcome	The measure and its definition
Stop or replace the objective	The intended outcome no longer serves the strategy	The allocation of effort
Run a focused test first	A critical assumption remains unresolved	The evidence available for the next commitment

Do not carry an OKR forward unchanged simply because work remains unfinished. Reconfirm its importance against other priorities.

If you revise a target, explain the new basis. Preserve the previous cycle’s result so the revision does not erase the miss. OKRs for SaaS Teams provides context for shaping the next cycle’s objectives and key results.

How do you turn the review into follow-through?

Write each decision in a form that can be checked:

We will change this action, because we observed this evidence. This owner will act by the agreed checkpoint. We will review this measure or observation before deciding whether to continue.

Keep the decision list small enough to resource. Remove or defer competing work where necessary.

Before closing, check that:

• Every action has an owner who accepts it.
• Each action addresses an identified cause or uncertainty.
• Shared dependencies have an explicit commitment.
• The next review has a clear evidence requirement.
• Unresolved questions are visible in the next cycle’s planning.

Which common mistakes should you avoid?

Lowering every missed target. First establish whether the target, approach, capacity or measurement needs to change.

Explaining the miss with effort. Acknowledge hard work, then examine what it produced and what constrained it.

Treating external change as a complete explanation. Review the response as well as the event. Sometimes the useful decision concerns earlier detection or contingency planning.

Rewarding confident stories. Keep plausible explanations separate from supported conclusions.

Using the meeting for personnel judgments. Review commitments and decisions openly; handle individual concerns through the appropriate separate conversation.

Waiting for the retrospective during an active problem. Address urgent customer, operational or safety issues immediately. Review the decisions after the situation is stable.

What else should teams know about missed OKRs?

Related reading

• OKRs for SaaS Teams — plan the next cycle with clearer objectives and key results.
• OKRs vs KPIs — distinguish intended changes from ongoing measures.
• OKR Examples for B2B Marketing — review examples when shaping marketing objectives.