- What OKRs actually are, and how they differ from KPIs and ordinary goals
- The right granularity for individual OKRs vs team OKRs
- A repeatable structure for writing objectives and key results
- Weekly, monthly, and quarterly cadences that keep OKRs alive
- Three real cases and the most common failure patterns
Most knowledge workers drift. Not because they are unmotivated, but because the work in front of them is louder than the direction they meant to be going. By Friday, the week was full; by the end of the quarter, the quarter was full; and somewhere along the way, the original intent got buried under whatever was most urgent. OKRs are one answer to that drift. They are not magic, and they do not replace execution. They are a structure for making sure the work that gets done is pointed where you wanted to go.
The reason OKRs feel corporate is that companies often deploy them badly — as a top-down compliance exercise rather than as a thinking tool. Used well by an individual, they are the opposite: a way to think clearly about what you are actually trying to do, and to check whether the work is doing it.
OKR is not KPI
The two are often confused, and the confusion wrecks both. A KPI (key performance indicator) measures something that is already happening — response time, defect rate, monthly revenue. It is a gauge. An OKR (objective and key results) describes what you are trying to change — the destination and the evidence you are moving toward it. It is a target.
| Dimension | KPI | OKR |
|---|---|---|
| What it is | A measurement | A goal |
| Time horizon | Ongoing | Quarterly |
| Direction | Maintains current state | Pushes toward new state |
| Completion target | 100% baseline | ~70% (stretch) |
Mixing them produces goals that are neither. “Keep response time under 2 hours” is a KPI dressed as an OKR. It describes maintenance, not change. A real OKR would be “reduce average response time from 4 hours to 1 hour by Q3” — a movement, not a status.
The right granularity for ICs
Company OKRs tend to be ambitious and broad. Individual OKRs fail when they copy that shape. An IC's OKRs should be small enough to actually be moved by one person in a quarter, and concrete enough that completion is unambiguous.
Three rules of thumb help:
- One to three objectives per quarter. More than three is a sign the OKRs are tasks dressed up as objectives.
- Two to four key results per objective. One is too fragile; five or more is usually a list.
- Each key result should be a number. If you cannot express it numerically, it is probably an objective, not a key result.
Writing objectives and key results
A good objective is qualitative, memorable, and slightly uncomfortable. It describes a state you want to reach, not a task you want to complete. Good key results are quantitative, time-bound, and evidence that the objective happened — not the work itself.
A bad OKR: “Improve the docs. KR1: write 10 articles. KR2: update 5 existing articles. KR3: review all docs.” This is a task list, not an OKR. The objective has no destination, and the key results measure effort, not outcome.
A better version: “Make the docs the first place users go for answers. KR1: reduce docs-related support tickets from 40/week to 15/week. KR2: increase doc search success rate from 55% to 75%. KR3: reach a 4.5/5 helpfulness rating on doc pages.” The objective describes a destination. The KRs measure whether you arrived.
Weekly, monthly, quarterly cadence
OKRs die without a cadence. The work of the quarter is done in the weeks, and if the weeks do not touch the OKRs, the quarter will not either. A simple three-layer cadence keeps them alive without making them the focus of every day.
| Cadence | Time | Focus |
|---|---|---|
| Weekly check | 10 min | Which OKR does this week's work advance? Any blockers? |
| Monthly review | 30 min | Are the KRs moving? Is the objective still right? |
| Quarterly retrospective | 1 hour | What did we learn? What changes for next quarter? |
The weekly check is the most important and the easiest to skip. Without it, the OKR sits in a document and only gets touched at the end of the quarter, when it is too late to do anything about it. The check does not need to be long — it needs to be honest.
Three real cases
Case 1: the engineer trying to grow beyond implementation
Objective: Become a technical leader on the payments team. KRs: (1) Lead design of two cross-team features, (2) Give one internal tech talk, (3) Mentor one junior engineer through their first major project. The KRs are evidence of leadership, not just activity.
Case 2: the writer building an audience
Objective: Establish authority in the async work space. KRs: (1) Grow newsletter subscribers from 800 to 2,500, (2) Publish 8 substantive articles, (3) Get cited or quoted in 3 external publications. The objective is directional; the KRs test whether the direction is producing real signal.
Case 3: the manager reducing their own bottlenecks
Objective: Stop being the bottleneck on team decisions. KRs: (1) Reduce my own decision turnaround from 3 days to 1 day, (2) Delegate 5 recurring decisions to the team, (3) Reduce my meeting hours by 25%. The KRs measure behavior change, not effort.
Common failure patterns
- Tasks dressed as objectives. “Ship the new dashboard” is a task, not an objective. The objective is the change the dashboard creates.
- Too many OKRs. Five objectives means no focus. Pick three and accept the rest will wait.
- KRs that measure effort, not outcome. “Write 10 articles” measures activity. “Reach 5,000 readers” measures outcome.
- No weekly check. OKRs without weekly contact die quietly by week six.
- Treating 100% as success. If you hit 100%, the OKR was too safe. 70% on a real stretch is healthier than 100% on a comfortable goal.
- OKRs describe change, not maintenance. KPIs measure maintenance. Do not confuse them.
- One to three objectives, two to four KRs each. More is usually hidden task lists.
- KRs measure outcome, not effort. If your KR is “do X,” rewrite it as “change Y.”
- Weekly contact is what keeps OKRs alive. Quarterly OKRs without weekly checks are decoration.
Frequently asked questions
Should individual contributors use OKRs or just goals?
Either works; the difference is mostly framing. OKRs force you to separate the aspirational objective from measurable key results, which tends to expose weak goals that are actually just tasks. If your existing goal-setting already does this, OKRs add little. If your goals tend to be vague or task-shaped, the OKR structure helps.
How many OKRs should an individual contributor have per quarter?
One to three. More than three is usually a sign that the OKRs are tasks dressed up as objectives. A useful test: if you cannot complete a quarter successfully while only working on these three, the OKRs are too many or too broad.
What is a healthy OKR completion rate?
About 70%. Consistently hitting 100% usually means the OKRs were too easy; consistently hitting 30% usually means they were either unrealistic or the work was redirected mid-quarter. The point is stretch, not perfection.