Product
Drafting OKRs from Jira, GitHub and your HR system with AI
How OKR.ai uses MCP to turn real projects and teams into a first draft worth editing.
·6 min read·Deep dive
Running your first OKR quarter, week by week
A 13-week calendar for setting, checking in on and grading OKRs without drowning in meetings.
Writing OKRs is the easy part. What decides whether they change anything is the rhythm around them: when they’re set, how often they’re looked at, and how the quarter closes. This playbook lays out a 13-week quarter, week by week, with the minimum set of rituals that make OKRs work without turning them into a meeting factory.
Key takeaways
- Weeks 1–2: set company OKRs and align teams in parallel. Week 3: acknowledge and approve, then lock.
- Weeks 4–12: a 15-minute weekly check-in. Update the number, give a confidence level, raise blockers.
- Week 7: a mid-quarter review to re-plan, not to rewrite goals.
- Week 13: grade on a 0.0–1.0 scale, run a short retro, and carry lessons into the next quarter.
- Start your first quarter with leadership plus one level. Expand once the rhythm sticks.
The 13-week OKR calendar
| Week | What happens | Who | Time |
|---|---|---|---|
| Before W1 | Leadership drafts 2–4 company objectives with baselines and targets | Exec team | Half a day |
| W1–2 | Function leads and managers draft aligned OKRs in parallel; check coverage and dependencies | Every manager | 1–2 hours each |
| W3 | Owners acknowledge their set; managers approve; OKRs are locked | Owners, managers | 30 minutes each |
| W4–12 | Weekly check-in: update numbers, confidence, blockers | Every team | 15 minutes a week |
| W7 | Mid-quarter review: what’s off track, what to change in the plan | Leadership, function leads | 60–90 minutes |
| W12 | Leadership starts drafting next quarter’s company OKRs | Exec team | A few hours |
| W13 | Final grading and a short retro per team | Every team | 30–45 minutes |
Weeks 1–2: set and align
The goal of the first two weeks is a complete, aligned set of OKRs, not a perfect one.
- Publish the company OKRs on day one. Teams can’t align to something they can’t see. Ideally leadership drafts them in the last week of the previous quarter.
- Draft in parallel. Function leads write their OKRs against the company set, and managers start as soon as their own lead’s draft is stable. Don’t wait for the whole level to finish.
- Check coverage. Every company key result should have owners below it, and additive targets should sum. See key results that add up.
- Surface dependencies early. If Sales needs a feature by week 6, Product needs to know in week 1, not week 9.
Week 3: acknowledge, approve and lock
Each owner confirms their set is something they can commit to (acknowledge), and their manager confirms it’s the right contribution (approve). Then the set is locked: from here, numbers change weekly but the OKRs themselves change only if the strategy changes.
Write down which OKRs are committed (must hit 100%) and which are aspirational (stretch). It changes how you read the grades in week 13.
Weeks 4–12: the 15-minute weekly check-in
This is the ritual that makes OKRs work. Keep it short, keep it to the numbers, and do it every week, including the weeks when nothing moved.
For each key result, the owner reports three things:
- The current value, against baseline and target. Attach evidence where it’s cheap: a dashboard link, a report, a screenshot.
- Confidence that it will land: on track, at risk or off track.
- What’s in the way, and who can help.
Discuss only the key results that changed confidence. Everything on track gets a nod and nothing more. If the check-in regularly runs past 15 minutes, a team has too many key results.
Week 7: the mid-quarter review
Halfway through, leadership and function leads look at the whole tree for 60–90 minutes. The question isn’t “are we hitting the numbers?” but “given what we now know, is the plan still right?”
- Move resources towards key results that matter most and are at risk.
- Kill or swap initiatives that aren’t moving their number.
- Change an OKR only if the strategy changed, for example a market shift, an acquisition or a major incident. “It’s hard” is not a reason.
Week 13: grade and retro
Grade every key result from 0.0 to 1.0, usually as progress from baseline to target. The objective’s grade is the average of its key results.
Then run a 30-minute retro per team with three questions:
- What did we learn about our customers, product or market?
- Were our targets well calibrated? Which ones were too safe or too wild, and why?
- What should we stop, start or continue in how we run OKRs?
Snapshot the final state before the next quarter begins, so you can compare quarters later.
Tips for your first OKR quarter
- Start small. Run the first quarter with leadership and one level below. Expand once the weekly rhythm sticks.
- Expect to over-reach. Most teams set too many OKRs and targets that are too high the first time. That’s fine; calibrate in the retro.
- Keep OKRs out of compensation. Tying bonuses to scores leads to sandbagged targets.
- Put the numbers where people already look. If OKRs live in a slide deck, they will be read once a quarter.
- Write good OKRs first. A great cadence can’t save a to-do list. Start with how to write good OKRs.
Automating the rhythm in OKR.ai
OKR.ai builds this cadence into the tool. Approvals are part of the workflow: owners acknowledge, managers approve, and anything changed since approval is flagged as new or changed. Weekly progress updates carry evidence files, and Suggest from context pre-fills values from Jira, Linear or your incident tool. The scheduler takes recurring snapshots of OKRs, progress and initiatives, so the end-of-quarter state is saved automatically.
Frequently asked questions
Should OKRs be quarterly or annual?
Most companies use annual company objectives with quarterly OKRs for teams. Quarters are long enough to move a real number and short enough to learn and adjust four times a year.
How long should OKR planning take?
Aim to have every team’s OKRs approved by the end of week 3. If planning takes a month, there are too many OKRs or too many approval layers.
Can we change OKRs mid-quarter?
Change the plan freely, but change the OKRs only when the strategy changes. Record what changed and why, so the end-of-quarter grade still means something.
Who should run the weekly check-in?
The team’s manager, using the numbers the key result owners have already updated. If the meeting is spent collecting numbers, ask owners to update before it starts.