Top down vs bottom up OKRs: which works better?
Neither extreme works: leadership sets a few company objectives and the why, teams propose their own key results and initiatives, and the two meet in a short negotiation. The healthy pattern blends clear direction with the judgement of the people closest to the work.
By Reuben Chircop, OKR specialist and leadership coach
Top down OKRs
In a top down approach, leadership writes the objectives and key results, then assigns them to teams. The direction is clear and the organisation can move quickly without waiting for every team to propose a plan.
That can help in a crisis, when a new strategy needs a decisive start, or in a small company where leadership is close enough to the work to understand the constraints. It gives teams an immediate priority when uncertainty would otherwise leave them pulling in different directions.
The problem begins when direction becomes prescription. Teams copy leadership key results into their own documents, even when they cannot directly influence those measures. There is a neat hierarchy of numbers, but no clear account of what each team will actually contribute.
Ownership then means accepting an instruction rather than making a commitment. Sandbagging, where people negotiate an easy target, is replaced by compliance: people agree to the number because arguing feels pointless. The target looks ambitious, but the honest conversation about capacity and risk never happens.
Bottom up OKRs
In a bottom up approach, teams propose objectives and key results from their knowledge of customers, systems and day-to-day work. Leadership reviews those proposals rather than starting with a prescribed set.
This helps when teams are mature, understand the strategy and can make sound trade-offs. They often see opportunities and constraints that a company-level view misses. Writing their own measures also makes it easier to explain why the work matters and to change initiatives when evidence changes.
Without clear company direction, however, each team can optimise its own corner. One team might prioritise faster delivery while another adds checks that slow the same process. Both can make progress locally without improving the outcome the company needs.
Another risk is too many objectives. If every worthwhile idea becomes an OKR, the organisation has a collection of local wish lists rather than a strategy. Bottom up input needs boundaries: shared priorities, limited capacity and permission to leave good ideas out of this cycle.
Top down, bottom up and the blend at a glance
| Aspect | Top down | Bottom up | The blend |
|---|---|---|---|
| Who writes objectives | Leadership writes and assigns them | Teams propose their own | Leadership sets company direction; teams define their contribution |
| Who writes key results | Leadership prescribes the measures | Teams choose the measures | Teams propose measures; both sides negotiate commitments |
| Alignment | Clear on paper, sometimes weak in practice | Depends on shared strategic context | Explicit links and trade-offs agreed together |
| Ownership | Can become acceptance of instructions | Strong local ownership | Team ownership within shared priorities |
| Speed | Quick to issue, slower if teams cannot deliver | Quick to draft locally, slower to reconcile | A prepared draft and one alignment conversation |
| Main risk | Compliance replaces honest judgement | Local optimisation replaces company progress | Negotiation drifts without a decision maker |
How the blend works in practice
Leadership owns the company direction. Teams own a proposal for their contribution. Neither draft is final until the two have checked the assumptions together.
- Leadership shares two to three company objectives and the context. Explain the why, the choices behind the strategy and the constraints. Teams need to know what matters and what is deliberately not a priority. Share draft company key results as something to discuss, not a finished list to cascade.
- Teams draft their own key results and initiatives. Ask what change the team can contribute, how it will measure that change and what work it thinks will move the measures. Keep initiatives separate from results so the team can change its approach without moving the goalposts.
- Hold one alignment conversation per team. Discuss what we will contribute and what we will not. Check capacity, dependencies and competing demands. Resolve overlaps with other teams and record the trade-offs, rather than adding another objective to keep everyone happy.
- Leadership adjusts company key results. Use what teams can commit to as evidence. If the combined contributions do not support the company target, revisit capacity, scope or the target itself. Do not quietly turn a strategic ambition into a promise nobody can keep.
- Agree the check-in rhythm. Name the outcome owners, decide when numbers and confidence will be updated, and put reviews in calendars. Use check-ins to surface learning and blockers, not to score individuals on attainment.
As a rule of thumb, aim for roughly 60% of team key results to be proposed by the team itself. This is not a statistic or a quota. It is a prompt to check whether teams have real room to shape the plan. The right balance depends on the situation; the important thing is that contributions are discussed rather than simply assigned.
Signs your balance is off
These signs are starting points for a conversation, not proof of one cause. Ask the teams what is happening before deciding how to fix it.
| Sign | Likely cause | Fix |
|---|---|---|
| Every team key result is a copy of a company one | Cascading has replaced team judgement | Ask each team for the specific contribution it can influence and measure |
| Teams have objectives nobody above them recognises | The strategy is unclear or alignment was skipped | Share the company context and agree which outcomes belong in this cycle |
| Planning takes more than three weeks | Too many approval rounds or unresolved priorities | Reduce the objective list and name a decision maker for the remaining trade-offs |
| Key results are all safe targets | Teams fear missing a target or being judged on attainment | Separate learning from individual evaluation and negotiate credible stretch openly |
FAQ
If drafting the first version is the slow part, the free OKR generator from OKRgen drafts one objective with three key results as a starting point for the team conversation.
Keep reading
What is an OKR?
The meaning, an example, and the three layers most guides skip.
OKR planning
When to start next year's OKRs, and a calendar that works.
Why OKRs fail
Five reasons implementations stall, and the order to fix them.
OKR vs KPI
The difference, and how they work together.
OKR template
A simple format you can copy, with filled examples.
OKRs and performance reviews
How to link reviews to evidence, not attainment scores.
Personal OKRs
A simple template and four examples for a season of change.
OKR glossary
Plain definitions of OKR terms and common mistakes to avoid.
OKR examples
Objectives and key results for every team.
Free OKR generator
Draft an objective and key results in seconds.