You know the method, all that is missing is the support. Here is a free OKR template in Excel format (compatible with Google Sheets and LibreOffice): two sheets ready to fill in, personal OKRs and team or startup OKRs, with concrete examples and progress computed for you. Direct download, no email required.
Download the OKR template (.xlsx, 9 KB) (the sheet labels are in French)
| Column | What goes in it | Why it exists |
|---|---|---|
| Objective | A qualitative direction, with no number | Merged across all the rows of its key results, so an objective with none is impossible to miss |
| Key result | A verifiable measure, with its unit | This is the only row you will update every week |
| Starting value | Where you stand on day 1 | Without it, 40 % progress means nothing |
| Target value | Where you want to be at the end of the cycle | Forces you to choose between ambition and fiction |
| Current value | Today's measure | The only number you touch during the weekly review |
| Progress | Nothing, it is a formula | (current minus start) divided by (target minus start), bounded to 0 and 100 % |
| Notes and next action | One sentence, not a journal | The most neglected and most useful column: it turns an observation into a decision |
Progress is deliberately bounded: a key result at 130 % does not exist in this template. Overshooting a target is not a performance, it is a sign the target was badly set, and letting that overshoot inflate the average would hide an objective falling behind right next to it.
This is the step that decides the value of the whole quarter, and it is the one people rush. Two symmetrical mistakes lie in wait.
The first is sandbagging: setting a target you already know you will hit. It produces a green sheet and a quarter with no effect. The test is simple: if you are 90 % certain of getting there, that is not a key result, it is a forecast.
The second is the decorative target: "go from 2,000 to 50,000 visits". Nobody believes it, so nobody uses it to make trade-offs, and the objective is quietly abandoned by week 4. A useful target is uncomfortable, not absurd.
The working benchmark: aim at a target you give yourself a 50 to 70 % chance of reaching. In OKR culture, finishing a cycle around 70 % is the sign of a well-calibrated objective, and 100 % everywhere means you aimed too low.
And if you do not know your starting value, do not invent it: spend the first week measuring it, and start the cycle with a real number. A quarter that begins on a false baseline produces twelve weeks of false conclusions.
The personal OKR sheet ships with this objective already entered, captured at week 4 of a 12-week cycle:
| Key result | Start | Target | Current | Progress |
|---|---|---|---|---|
| Run 3 sessions a week for 12 weeks (36 sessions) | 0 | 36 | 9 | 25 % |
| Go from 0 to 10 km without stopping | 0 | 10 | 4 | 40 % |
| Sleep 7 hours 30 on average per night (hours) | 6.5 | 7.5 | 7.0 | 50 % |
Three things are worth noticing. First, the unit is in the wording (sessions, km, hours): without it, after a month you no longer know whether "9" is a number of runs or of kilometres. Second, the third key result starts at 6.5 rather than zero, which is exactly the case where progress computed without a starting value would lie, showing 93 % instead of 50 %. Third, the first row is at 25 % in week 4 of 12, which is precisely on schedule: the comparison that counts is not "where am I", but "where am I relative to how much of the cycle has elapsed".
That is the reading to get into the habit of, and it is the same one as a productivity dashboard: what steers a decision is the trajectory, never the absolute value.
The template is only worth as much as this ritual. Without it you will have filled in a nice table in January and reopened it in April, with the unpleasant feeling of having lost a quarter.
Put a recurring 10-minute slot in the calendar, on the same day each week, and block it like an appointment: it is the most profitable use of time-blocking there is. During those 10 minutes, three moves, in this order:
That is all. The review is neither a report nor a tribunal, it is a set of points. It has done its job if it produces one line in your following week.
At the end of the quarter, the average of an objective's progress gives its score, between 0 and 100 %. The reading grid, inherited from practice at Intel and then Google:
The score is there to calibrate the next cycle, never to evaluate a person. As soon as an OKR score decides a bonus, targets collapse within three quarters: everybody learns to promise only what they already know they can deliver. And if your need is to monitor rather than to change, what you want is a KPI, not an OKR. The distinction is set out in OKRs vs SMART vs KPIs.
The team and startup sheet is enough for 2 to 5 people. Three adjustments make it usable collectively:
If a term in the field escapes you (key result, cycle, KPI, sandbagging), the productivity glossary defines them all in one sentence.
This template is the best way to start today. Let us be honest about its limits, familiar to anybody who has tracked OKRs in Excel:
If those limits are holding you back, Cosmo has OKRs built in: progress updated as you act, a chart of key results achieved on the dashboard, and your goals living next to your scheduled tasks and your habits. Free, with a demo and no sign-up.
Yes: import the .xlsx file into Google Sheets (File then Import) and the progress formulas are preserved. It is also compatible with LibreOffice Calc and Numbers.
Yes, freely. If each member tracks their own OKRs, duplicate the team and startup sheet once per person, keeping a single file for everybody rather than one file each.
No. The link is direct, with no form and no sign-up. We would rather you judged the content than gave us your address.
Two to four per cycle, with 2 to 5 key results each. The template offers more for layout comfort, not as a quota to fill. Beyond four, the weekly review runs past ten minutes and stops happening.
Yes, and it is better than waiting. Take the remaining weeks as your cycle length and scale the targets down in proportion: a well-tracked seven-week cycle beats a twelve-week cycle that starts in a month.
Pick up at the next one, without catching up the missed weeks. Nothing in the template depends on history, only the current value counts. It is mostly a signal that the review slot needs moving, not that willpower needs raising.
← All articles · Try Cosmo for free
User guide · Cosmo for companies · Blog · For freelancers · For students · For managers · For teams · About · Free sign-up · Sign in · Legal notice · Privacy · Terms · contact@thecosmo.app