Everything you need on one page: the shape of a good OKR, the rules that keep it honest, and three worked examples you can copy. New to OKRs? Read OKR vs KPI first. About 8 minutes to read.
| ❌ Output (a task, goes under initiatives) | ✅ Outcome (a Key Result) |
|---|---|
| Launch a new onboarding email series | Raise users active in week 1 from 30% to 45% |
| Run 4 security training sessions | Cut phishing-test click rate from 18% to 5% |
| Hire 2 support agents | Cut first reply time from 9 hours to 2 hours |
Score each Key Result from 0.0 to 1.0 by how far you moved from X towards Y. Average them for the Objective.
Committed vs stretch: some goals must hit 1.0, such as a legal deadline. Label those "committed" so nobody reads 0.7 as fine. The compliance example below is one.
Staying compliant day to day is a KPI: "zero overdue access reviews". But when you need to reach a new state, such as passing an audit, gaining a certification, or closing a backlog of findings, that is a change with a deadline. That makes it a good OKR.
Why this works: KR1 and KR4 prove the destination. KR2 and KR3 prove the change will last after the auditors leave, which is the real point of the Objective.
Acquisition means winning new customers. It is easiest to think about as a funnel: many people hear of you, fewer try you, fewer still pay. Good acquisition OKRs pick the stage of the funnel where people drop out the most.
| Term | What it means |
|---|---|
| Conversion rate | The share of people who move from one stage of the funnel to the next. |
| CAC (customer acquisition cost) | Total sales and marketing spend ÷ new customers won. $8,800 ÷ 40 = $220. |
| Activation | The moment a new user gets real value for the first time, such as a clinic's first booked appointment. |
| Churn | Customers who stop paying in a period, as a share of all customers. |
These are the two most common ways to put a number on how customers feel. They sound alike but measure different things: CSAT measures one moment, NPS measures the whole relationship.
The question: "How satisfied were you with [this support chat / your delivery / this visit]?" Customers answer from 1 (very unsatisfied) to 5 (very satisfied), usually right after the moment happens.
The score: the percentage of people who answered 4 or 5.
The question: "How likely are you to recommend us to a friend or colleague?" Customers answer from 0 to 10. It is usually asked every few months, not after every interaction.
The score: answers fall into three groups. NPS = % Promoters − % Detractors. Passives count towards the total but are not added or taken away. The result ranges from −100 to +100. It is a plain number, not a percentage.
| CSAT | NPS | |
|---|---|---|
| Measures | One moment or interaction | The overall relationship and loyalty |
| Question | "How satisfied were you with …?" | "How likely are you to recommend us?" |
| Scale | 1 to 5 | 0 to 10 |
| Score | % of 4s and 5s (0% to 100%) | % promoters − % detractors (−100 to +100) |
| When to ask | Right after the interaction | Every few months |
| Moves | Fast, within weeks | Slowly, over months |
| Best for | Fixing a specific experience | Tracking whether customers would back you |
Be careful with small numbers. With 40 survey answers, a few people can swing NPS by 10 points. Always show how many people answered next to the score. And compare NPS with your own past results or your industry, not with companies in very different markets.
Why use both? CSAT moves quickly, so it tells you within weeks if your changes are working. NPS moves slowly but tells you whether those changes add up to loyalty. A rising CSAT with a flat NPS often means you fixed the small things but not the big one.