A clean number from recent OKR research puts a price on something operators have suspected for years: teams with single-owner key results complete 26% more of their goals than teams that share ownership. The same research finds that only 46% of teams can honestly say every OKR has a clear owner.
The first number is the headline. The second number is the more honest one. More than half of all OKRs in the dataset don't have a person whose week is structured around them — and you can predict, just from that fact, which quarter is going to close green and which is going to quietly slip.
What "shared" actually means in practice
When people say a KR is "shared," they almost always mean one of three things, none of which produce ownership in any operational sense.
The duo. Two names on the KR. Each privately assumes the other will drive the weekly cadence. Neither does. By week four, when somebody finally asks, both will explain, with complete sincerity, that they thought the other one was leading it. They are both telling the truth. That's the problem.
The function. "Product owns this" — which means a department of fifteen people, none of whom can be paged when the KR slips. The function is responsible in the same way that "society" is responsible for things. Diffuse, plural, untouchable. If your KR's owner is a noun rather than a name, it's unowned.
The committee. A working group with a steering doc, a Slack channel, biweekly syncs, and not a single person whose performance review depends on the result. Committees are excellent at producing other committees. They are terrible at producing outcomes, because nobody's individual calendar is reorganised around the work.
All three feel responsible from the inside. None of them produce the weekly forcing function that single ownership produces almost automatically. That's why the gap is 26 points and not 5.
Why single ownership compounds
A single owner gets paged when the number moves the wrong way. A single owner reorganises their week around the KR — they choose what to deprioritise, what meetings to skip, which dependencies to escalate. A single owner notices, on a Wednesday morning, that a dependency two teams away is slipping, and they personally walk over and renegotiate it. None of those reactions happen for a shared KR, because no individual's calendar is built around it. The KR doesn't make it onto anyone's "what is this week about?" list, which means it doesn't make it onto anyone's actual week.
Multiply this dynamic across a quarter of KRs and the 26-point gap stops looking mysterious. It's not a culture problem. It's a calendar problem. Accountability is not a feeling. It's whose week changes when the number changes.
The objection — and the actual answer
The most common pushback is reasonable on its face: "But our work genuinely requires cross-functional collaboration. The KR really does depend on three teams. Naming one owner misrepresents how the work actually has to happen."
Yes — and the way to express that truthfully is not shared ownership. It's a single accountable owner with explicit, named contributors and a documented dependency model. The owner is accountable for the outcome. The contributors are accountable for their specific contribution. Both relationships are visible and queryable. Neither is shared.
This distinction matters in the planning artefact. "Maria owns the activation KR. Engineering contributes 30% of two named engineers' bandwidth for six weeks. Marketing contributes the lifecycle automation work. The dependency on the data platform is on the critical path." That paragraph allocates real responsibility. "Product, Engineering and Marketing co-own activation" does not allocate anything — it just enumerates the rooms the work could theoretically live in.
Why companies keep choosing shared ownership anyway
If the data is this stark, why do most leadership teams keep distributing ownership across functions? Because naming a single owner forces an uncomfortable conversation. It forces the leadership team to decide, in public, which function's leader is going to be subordinate to which other function's leader for the duration of a particular bet. That conversation is exactly what cross-functional ownership is meant to avoid — and it is exactly the conversation that has to happen for the bet to actually move.
Shared ownership is the planning artefact you reach for when the leadership team isn't willing to have the harder conversation. Which means it's the planning artefact that's most predictive of the bet not happening.
A practical reframing
If you're staring at a quarter of KRs with shared ownership, you don't need to relitigate the strategy. You need to do one thing for each KR: name a single accountable owner, write down the named contributors with their explicit percentage of week, and document the dependencies. The KR's strategic content doesn't change. The probability of completion changes by roughly 26 points.
The Vindaris view
Every KR should have exactly one accountable owner. Contributors should be explicit and separate. Dependencies should be first-class entities, not implications. When the system refuses to model a result as "co-owned," the data becomes unambiguous: co-owned is a polite name for unowned. Put strategy, capacity and ownership on one surface and the 26-point gap stops being a research finding and starts being a property of how the company is structurally built to operate.