How to
How to see who hasn’t written OKRs
Short answer: Put every person on the board. Show Has, Missing, Draft, and Pending for the active cycle. Empty coverage is the signal. Do not omit people with nothing written to “reduce noise.”
That is not a reporting preference. It is how you know whether the company actually ran OKRs this quarter.
The request that sounds reasonable
A manager asks to filter the directory to “people with OKRs.” An admin wants a cleaner board for the exec review. Someone new joined mid-cycle; someone is on leave; a contractor “doesn’t do OKRs.” The spreadsheet that used to list twenty names now lists twelve, and the twelve look committed.
You did not fix coverage. You hid the denominator.
If the question is “who hasn’t committed this quarter?”, the only honest answer includes the people with a blank. A list of writers is a highlight reel.
What the research says
Goal-setting only works when the goal is specific, difficult, and visible enough to get feedback. That is not a slogan. It is one of the most replicated findings in organizational psychology. Locke and Latham’s theory, built from decades of studies and several meta-analyses, is blunt on the moderators: specific challenging goals beat “do your best,” and goals plus feedback beat either one alone. People adjust effort and strategy when they can see progress. They cannot do that for a goal that was never written down, and the company cannot give feedback on a person it quietly dropped from the list.
So the first failure is not a low score. It is an unwritten goal. There is nothing to be specific about, nothing to attach feedback to, and nothing for a peer to align to.
Industry surveys keep finding the same hole one layer up. A 2026 benchmark of 200+ operators reported that 65% of teams say their OKRs are not tied to company goals, and that half of key results have no named owner. Those are coverage problems wearing strategy clothing. You cannot cascade into a blank. You cannot assign a single owner to a key result that does not exist. The same survey found teams that launch in under a week complete more goals than teams that spend the quarter designing the perfect system — which is another way of saying start with the honest empty board, then fill it.
Abandonment follows the same pattern. When programs die, operators cite low engagement and no clear ownership more often than “the framework was wrong.” People disengage from a process that cannot show who is in and who is out. A directory that only shows writers teaches the org that silence is a valid strategy.
Psychological safety is the usual objection: “If we show Missing, people will feel shamed.” Safety is the ability to tell the truth in public, not the ability to disappear from the roster. Amy Edmondson’s work on teaming is about making it speakable when work is incomplete. A loud Missing chip is that conversation, scheduled. A filtered-out row is the opposite — the gap becomes undiscussable, which is how polite companies fail the method.
Why hiding the empty seats breaks the cascade
Company objectives are supposed to be the thing team objectives align to. Team objectives are supposed to be the thing individual objectives align to. That graph is a lie if a third of the nodes were never drawn and the UI pretends they were never supposed to be.
Three things happen when empty people fall off the board:
- The review gets dumber. Leaders grade the people who wrote something. The people who did not write something stop being a topic. The meeting becomes a tour of the committed, not a picture of the company.
- Alignment has nowhere to land. An individual who might have parented to a team objective never gets asked, because they are not on the list. Unaligned work and missing work are different failures. You need both visible.
- Ownership dissolves. Shared “we’ll get to it” is how key results end up with no named owner. Coverage that is owner-keyed — one person, live objective or live key result — makes that dodge expensive.
Contributors and Watchers are useful for “who else is involved?” They are not coverage. If helpers could clear Missing, every popular objective would manufacture a full board.
How to run coverage in the open
1. Keep the whole directory
Every authenticated member of the org can already read every objective. The People surface should show every person for the selected cycle (default = active). Zero-OKR people stay on the page. Teams with zero members or zero team-level OKRs stay on the Teams lens too.
2. Use four states, not a progress ring
| State | What it means |
|---|---|
| Has | Owns a live objective or a live key result this cycle |
| Missing | Owns neither |
| Draft | Work is started and not yet live |
| Pending | A write is waiting on manager or admin approval — still org-visible |
Do not fold Draft or Pending into Has. Do not fold Missing into “not applicable.” A progress ring that averages the writers is a costume.
3. Count live work, not history
Ended or cancelled work does not clear Missing. Display counts may include history so the timeline stays honest. Status does not. The question coverage answers is “who is on the hook now?”
4. Let a key result owner count
If you own a live key result on someone else’s objective, you are covered. That is still one accountable person. It is not a team score and it is not a helper list.
5. Walk the gaps in the review
The biweekly agenda should include Missing as a first-class section, not an appendix. Ask “who still has nothing written?” before you admire scores. Scores on a partial roster are not a company result.
6. Write something small rather than waiting for perfect
The research on launch speed is the practical move: a draft objective this week beats a polished cascade in week ten. Missing is supposed to get quieter because people write, not because the filter got stricter.
What not to do
Do not ship a “people with OKRs” default. That is a highlight reel. Make the full roster the default. Filters can exist for finding a name. They must not become the exec view.
Do not let Contributors, Watchers, or a team label clear Missing. Involvement is not ownership. Team coverage is additive and never a permission wall.
Do not paper over new hires and leave. If someone is out of the cycle, say so where everyone can read it. Do not delete the row.
Do not treat empty Teams as a cleanup task. A team with no team-level OKRs is a gap on the team lens. Hiding it recreates the same lie one level up.
A worked example
Northstar Labs opens Q3. The directory has 24 people.
Week one: 6 Has, 3 Draft, 15 Missing. The review is short and uncomfortable. That is the point.
Week three: 14 Has, 4 Draft, 2 Pending, 4 Missing. Two of the four Missing are on leave and labelled that way. Two have not written. The company can name them. Nobody has to guess from Slack silence.
If the admin had filtered to “Has + Draft” in week one, the exec packet would have said “9 people, looking good.” The 15 blanks would have become next quarter’s surprise.
If Missing feels too loud
Then the process is working.
A company that cannot stand to see who has not committed will not stand to see who missed a committed 1.0, or who has been Unaligned for two cycles. The directory is the cheapest honesty you will get. Gahmbit will not offer a quieter board. We would rather lose the “reduce noise” checkbox than ship a product that congratulates a partial roster.
Sources
- Locke, E. A., & Latham, G. P. (2002). “Building a Practically Useful Theory of Goal Setting and Task Motivation.” American Psychologist. Specific, difficult goals outperform vague ones; feedback is a critical moderator.
- Locke, E. A., & Latham, G. P. (2019). “The Development of Goal Setting Theory.” Goals plus feedback outperform either alone.
- Edmondson, A. The Fearless Organization and related teaming research: psychological safety is speakable incompleteness, not disappearance from the roster.
- OKRs Tool (2026). OKR Benchmark Report — 200+ operators; 65% of teams report OKRs not tied to company goals; ~50% of key results lack a named owner. Also summarized in Yahoo Finance.
- Related: Should different teams have different OKR cycles?
FAQ
- Does owning a key result on someone else’s objective count as coverage?
- Yes. Coverage is owner-keyed. A live key result you own still counts, even when the parent objective belongs to someone else. Contributors and Watchers never clear Missing.
- What if someone is drafting and not ready to publish?
- That is Draft — a visible state, not an empty seat. Pending is the same idea when a write is waiting on approval. Neither should be folded into Has or dropped from the list.
- Should we hide people who are new or on leave?
- Do not drop them from the directory to make the board look full. If they are out of the cycle, say so in the open. Silence is how Missing stops meaning anything.
- Do team OKRs clear individual Missing?
- No. Teams are an additive lens. A team with no team-level OKRs stays visible too. Person coverage remains primary.
- Is this free?
- Yes. The People directory and coverage chips are free core. No credit card. Optional AI is prepaid org credits only when a model runs.