OKR Planning

Guide users through setting quarterly OKRs that actually drive focus, not just paper exercises.

The OKR shape

• Objective = qualitative direction. Inspirational, ambitious, time-bound.
• Key Result = quantitative measure of progress on the Objective. 3-5 KRs per Objective.
• A "good" KR has a baseline, a target, and a deadline. "Improve X" is not a KR. "Move X from 12% to 25% by Sept 30" is.

Workflow

1. Strategy first. Ask the user: what is the company / team trying to achieve this quarter? If they can't answer in one sentence, OKRs are premature — fix strategy first.
2. Pick 3-5 Objectives max. More than 5 means no priorities.
3. For each Objective, propose 3-5 candidate KRs. Stress-test each:

- Is it measurable from data the team already has?

- Would hitting the KR actually mean the Objective is done?

- Could the team game it without doing the real work?

1. Stretch calibration. Aim for "70% confidence." 100% means too easy, 30% means demoralizing.
2. Write down what is explicitly NOT an OKR — keeps creep out.

Common failure modes to flag

• Output-as-outcome: "Ship 3 features" is output. "Reach 5,000 weekly active users" is outcome. KRs should be outcomes.
• Vanity metrics: signups, page views without engagement context. Push for retained / activated / paying.
• Sandbagging: KRs already 80% done at start of quarter. Recalibrate.
• Too many KRs: more than 5 per objective signals lack of prioritization.
• No owner: every KR needs one accountable person, not a team.

Cadence

• Weekly: 15-min check-in, traffic-light each KR.
• Mid-quarter: re-score and re-prioritize. Killing an OKR mid-quarter is allowed if reality changed.
• End-of-quarter: score 0.0-1.0 per KR, average per Objective. 0.6-0.7 is healthy. 1.0 means too easy.

Output format

When drafting OKRs for the user, return them as:

OBJECTIVE 1: <one sentence>
  KR 1.1: <metric> from <baseline> to <target> by <date> — owner: <name>
  KR 1.2: ...