okr template
OKRs are not a reporting tool. They are a thinking tool. If your team can hit every key result by doing exactly what they were already going to do, you have not written OKRs — you have written a status report.
This is the template. Copy the markdown block, paste it into Notion or Confluence, and fill in your numbers. What follows is the part most OKR guides skip: why the format works the way it does, where teams consistently break it, and how to run the scoring conversation without it becoming political.
The template
Copy everything between the horizontal rules below.
# OKRs: [Team / Squad Name] · [Quarter, e.g. Q2 2025]
**Owner:** [PM / Lead name]
**Created:** [Date]
**Review cadence:** [Weekly check-in / Bi-weekly]
**Last scored:** [Date — fill in at end of quarter]
---
## Objective 1: [Aspirational, qualitative goal]
*One sentence. No numbers. Should be uncomfortable if you squint at it.*
---
### KR 1.1: [Metric] from [baseline] to [target] by [date]
- **Baseline (now):** [Current number — go measure it before you write this]
- **Target:** [Stretch goal — 70% attainment = success; 100% means you aimed too low]
- **Data source:** [Where this number lives: dashboard name, SQL query, analytics tool]
- **Owner:** [Single person accountable for tracking this]
- **Mid-quarter check-in:** [Date + early signal you will look for]
### KR 1.2: [Metric] from [baseline] to [target] by [date]
- **Baseline (now):**
- **Target:**
- **Data source:**
- **Owner:**
- **Mid-quarter check-in:**
### KR 1.3: [Metric] from [baseline] to [target] by [date]
- **Baseline (now):**
- **Target:**
- **Data source:**
- **Owner:**
- **Mid-quarter check-in:**
**End-of-quarter score (0.0–1.0):** ___
**What moved the needle:** [Fill at end of quarter]
**What we learned:** [Fill at end of quarter]
---
## Objective 2: [Aspirational, qualitative goal]
### KR 2.1: [Metric] from [baseline] to [target] by [date]
- **Baseline (now):**
- **Target:**
- **Data source:**
- **Owner:**
- **Mid-quarter check-in:**
### KR 2.2: [Metric] from [baseline] to [target] by [date]
- **Baseline (now):**
- **Target:**
- **Data source:**
- **Owner:**
- **Mid-quarter check-in:**
### KR 2.3: [Metric] from [baseline] to [target] by [date]
- **Baseline (now):**
- **Target:**
- **Data source:**
- **Owner:**
- **Mid-quarter check-in:**
**End-of-quarter score (0.0–1.0):** ___
**What moved the needle:**
**What we learned:**
---
## Initiatives (how we plan to move the KRs)
*These are not KRs. They are bets. If they stop moving the metrics, drop them.*
| Initiative | KRs it affects | Owner | Status |
|------------|---------------|-------|--------|
| [Feature / experiment name] | KR 1.1, KR 1.2 | [Name] | In progress |
| [Feature / experiment name] | KR 2.1 | [Name] | Not started |
---
## What we are explicitly not doing this quarter
- [Work that was requested and deprioritised — name it so the conversation happened]
- [Work that is good but not connected to these objectives]
---
## Quarter-end retrospective
**Overall score:** ___
**Biggest contributor to results:**
**Biggest miss and why:**
**What we would do differently:**
**Carry-over to next quarter:**How to fill each section
The Objective
An Objective is a sentence, not a metric. It should tell a story about where the team is going, not how far it has traveled. “Become the most trusted payments experience for small merchants in India” is an Objective. “Increase D30 retention by 12%” is a Key Result.
The test for a good Objective: read it to a new engineer on their first day. Does it tell them what the team is trying to accomplish in human terms? If yes, it is an Objective. If it requires a data dictionary to interpret, it is a KR in disguise.
Keep at most two Objectives per quarter per team. Three is a stretch. Four means you are describing everything you plan to do, which is not strategy — it is a task list.
Red flags:
- The word “maintain” in an Objective. Maintenance is operations, not direction.
- Two Objectives that do not connect. If a team owns both “Grow the supply side” and “Improve checkout conversion,” you should be asking whether these belong to one team or two.
- An Objective that describes an output: “Ship the new onboarding flow.” That is a project milestone, not a strategic direction.
Key Results
Three rules that are non-negotiable:
1. Every KR needs a baseline before you write the target.
Most teams skip this. They write “increase retention from 40% to 55%” without checking what retention actually is today. When the quarter ends, they discover the baseline was 48%, which makes the target meaningless. Go pull the number first. If you cannot find it, your first KR is “instrument and establish baseline for [metric] by [date].”
2. Target 70% attainment as success, not 100%.
This comes directly from how Google runs OKRs, and it is consistently the most misunderstood part of the system. A score of 1.0 (full attainment) means you aimed too low. A score of 0.4 means either you aimed too high or the work was wrong. The sweet spot — where learning happens — is 0.6 to 0.75. This only works if your leadership actually believes it. If your manager treats a 0.7 as failure, your team will write sandbagged targets every quarter and you will have achieved nothing except a false sense of progress.
3. One data source, named upfront.
“We will measure this in analytics” is not a data source. “Active users as reported in Mixpanel, event: session_start, property: user_type = paid” is a data source. Name it when you write the KR. This prevents the end-of-quarter argument about whether you hit the target, which is the most demoralizing conversation a team can have after working hard for three months.
Initiatives
Initiatives are the work. They are not the goal.
This is the biggest structural mistake teams make with OKRs: they list their planned features as Key Results. “Launch referral program” is not a Key Result — it is an Initiative. The Key Result is “grow organic signups (referral channel) from 12% to 25% of new users.” The referral program is one possible way to move that number. If you launch it and it does not work, you have failed the KR even if the Initiative shipped.
The separation matters because it forces the team to ask: “Is this Initiative actually moving the metric?” If you are halfway through Q2 and the referral program launched but organic signups are still at 11%, the question is no longer “are we on track” — it is “do we keep betting on this Initiative or try something else?” That is the conversation OKRs are designed to create. It only happens if KRs and Initiatives are written separately.
What we are explicitly not doing
Write this section. Most teams omit it because it feels confrontational. That is exactly why it matters.
Every quarter, someone requests work that your team does not prioritise. If that work goes unacknowledged, the requestor assumes it is in your plan. When Q2 ends and they ask why it was not built, the team looks like they forgot. When it is listed explicitly — “re-platforming the notifications service: deprioritised in favour of retention work, revisit Q3” — the conversation happened. The requestor knows they were heard. The team is not blamed for forgetting.
The four anti-patterns
1. KRs that measure outputs instead of outcomes
Output: “Ship 3 new onboarding screens.” Outcome: “Increase onboarding completion rate from 54% to 72%.”
Outputs are easy to hit. You ship the screens on the last day of the quarter regardless of whether they worked. Outcomes force you to keep iterating until the metric moves.
2. KRs with no baseline
“Double our NPS” is not a KR if you have not measured NPS yet. “Establish NPS baseline (target: first measurement by end of week 3) and improve it by 8 points by end of quarter” is a KR. Do not let the absence of a baseline stop you from writing a KR — make establishing the baseline the first KR.
3. Too many KRs
More than three KRs per Objective is usually a symptom of one of two problems: either the Objective is too broad (it should be two Objectives), or the team has not made trade-offs and is hedging by listing everything. Three KRs is the maximum before the system loses focus.
4. OKRs written by one person
OKRs that are handed down from a manager to a team are not OKRs — they are targets with extra steps. The system only generates commitment when the people doing the work had a say in setting the goals. The PM’s job is to set the direction (the Objective) and negotiate the targets (the KRs) collaboratively with the team. That negotiation — where an engineer says “I think 72% is achievable but 80% requires work we have not scoped yet” — is where alignment gets built. Skip it and you get compliance, not ownership.
The scoring conversation
At the end of the quarter, run this in a 45-minute retrospective with the team, not as a solo PM exercise:
Step 1: Score the KRs first (10 min) For each KR, divide actual attainment by target. A KR targeting 50 new enterprise trials where you got 38 scores as 0.76. Do this together, not beforehand.
Step 2: Explain the score before judging it (15 min) For every KR below 0.5, the team answers: was the target wrong, was the work wrong, or did something external change? These are three different problems with three different responses. A wrong target means better calibration next quarter. Wrong work means the Initiative needed to change mid-quarter — why did it not? External change is the only one that requires no team learning.
Step 3: Separate score from performance review (5 min) Say this explicitly at the start of the conversation: “A score of 0.6 on a stretch goal is a good quarter. We are not grading people here. We are grading our aim.” If your company uses OKR scores in performance reviews, you will get sandbagged targets. The two systems are incompatible and leadership needs to choose.
Step 4: Write the learnings (10 min) Fill in the “What we learned” field for each Objective before the session ends. If you defer this, it never gets written. These notes become your most valuable input for next quarter’s Objective-setting.
Step 5: Carry-over and close (5 min) Explicitly name what carries forward and what does not. Close the quarter. Starting fresh matters psychologically — teams that never “finish” a quarter end up with objectives that drag on for six months with no accountability.
If your team ran OKRs last quarter — even informally — pull them up now.
For each Key Result:
- Find the actual ending number. Not an estimate. The actual number from your analytics tool, CRM, or dashboard.
- Divide actual by target. Write the score as a decimal (0.0–1.0).
- For each KR below 0.6, answer these three questions in writing: Was the baseline wrong? Was the target too aggressive? Did the Initiative fail to move the metric?
- For each KR above 0.9, ask: did we aim too low? What would we have targeted if we had been braver?
If you did not run formal OKRs, pick one metric your team is currently responsible for — retention, conversion, activation rate, whatever. Write a KR for it right now using the template format above. That means: finding the baseline (today’s number), setting a target (what good looks like in 90 days), naming the data source, and naming the owner.
You cannot write a good OKR for a metric you do not currently measure. If you discover you cannot find the baseline, write that down. That gap is more valuable information than a clean template.
It is OKR-setting week for Q3. You are a PM for a B2B SaaS product. Your team's activation rate — the percentage of trial accounts that complete the first core workflow — is currently 31%. Your VP of Product asks the team to set an activation OKR. Your lead engineer, Priya, suggests targeting 38%. Your growth analyst thinks 52% is achievable based on funnel analysis. Your VP privately tells you she wants to see 55% to 'show the board we're serious'. You have one hour to align the team on a number.
You have heard three different numbers from three different stakeholders. The board presentation is in six weeks. Priya is the one who will do most of the implementation work.
your path
Related pages
- Measuring Outcomes — how to find and defend baselines, and which metrics belong in OKRs vs dashboards
- Prioritisation Frameworks — when your OKR and your backlog are pulling in different directions
- PRD Template — how Success Metrics in a PRD connect to KRs in your OKR
- Stakeholder Management — running the OKR-setting conversation when leadership and engineering cannot agree on targets