A hoshin plan isn’t finished when everyone agrees with it. It’s finished when the first monthly review could run from it. A lock gate checks that before the year starts, rather than in the first awkward review of January.
In this guide
Key Takeaways
- →Agreement is not readiness. The test is whether January’s review could run from the plan.
- →Thirteen pass-or-fail checks across structure, correlations, people, agreement and operation.
- →Locking doesn’t freeze the plan. It changes how the plan changes.
- →Countermeasures, tasks and team key results still move freely after lock.
- →Lock what is ready and list the open items with dates. Don’t hold the whole plan for one escalation.
- →The record of what changed after lock is what makes next year’s reflection possible.
Agreed is not the same as ready
Planning seasons usually end with a sense of agreement. The matrix has been debated, catchball has closed, the executive team has signed off. Then the first review of the year opens and the gaps appear: KPIs with no baseline, priorities with two owners, plan values that are just the annual target divided by twelve, nobody sure when check-ins are due.
Each gap is small. Together they mean the first quarter is spent finishing the plan rather than running it.
The lock gate
Run this checklist in week 11 of the planning calendar. Every item is pass or fail.
Structure
- Every annual objective links to a breakthrough.
- Every priority links to an annual objective.
- Every KPI has a complete card: definition, formula, baseline, target, source.
Correlations
- Every primary correlation has passed the hypothesis gate: mechanism, contribution, lag window, kill signal.
- Every KPI has at least one primary correlation.
- Primary contributions fit each KPI’s gap.
People
- Every row has one named owner.
- The capacity check is resolved, with no team or person left overloaded.
- The stop-doing list is complete, with owners and dates.
Agreement
- Catchball is closed on every objective, or its escalation has been decided.
- Site targets add up to group targets where they should.
Operation
- Monthly plan values are phased by path shape.
- The check-in cadence and review calendar are set, with first check-ins dated.
What locking means
Locking doesn’t freeze the plan. It changes how the plan changes. Before lock, objectives, targets and correlations can be edited freely. After lock, changes to them go through the quarterly review, with a date and a reason recorded. That record is what makes next year’s reflection possible.
| Changes freely after lock | Changes through the quarterly review |
|---|---|
| Countermeasures and tasks | Targets |
| Check-in comments | Correlation grades |
| Team-level key results beneath priorities | Adding or retiring priorities |
| Meeting schedules | Owners of objectives and KPIs |
A plan that can be changed by anyone at any time isn’t locked. A plan that can’t be changed at all isn’t managed. Lock is the line between the two.
Partial lock
If a few items genuinely can’t close in time, such as a target waiting on a budget decision, lock everything else and list the open items with dates. Don’t hold the whole plan for them. The rest of the organization can start running its part of the plan in January while the open items close.
A worked example: a lock gate result
| Check | Result | Resolution in week 11 |
|---|---|---|
| Links to breakthroughs and objectives | Pass | |
| KPI cards complete | Fail: two baselines missing | Estimated from Q3 data, marked for confirmation in January |
| Primary correlations gated | Pass | |
| One owner per row | Fail: one shared priority | Assigned to the commercial director |
| Capacity resolved | Pass, after one priority sequenced to H2 | |
| Stop-doing list | Pass | |
| Catchball closed | Fail: one escalation pending | Partial lock; decision due 18 December |
| Site roll-up | Pass | |
| Plan values phased | Pass | |
| Cadence set | Pass; first check-ins due 30 January |
Illustrative.
Three fails, all resolved or contained in the week. The plan locked with one listed open item rather than waiting for it. The first January review ran from the locked plan, with the one pending target marked as such.
What the gate needs from the software
Thirteen checks, run in one week, across every row of the matrix. Done by hand that is a spreadsheet someone builds in week ten and nobody trusts by week twelve. Most of the checks are queries: things with no owner, KPIs with no baseline, correlations missing a kill signal, rows whose plan value is the annual target over twelve.
The distinction worth asking about is what lock actually does. In a document, lock is a convention: everyone agrees to stop editing. In a system it can be a state, where some fields move freely and others require a recorded reason, which is exactly the table above.
| Gate area | What the tool has to answer | How to test it in a demo |
|---|---|---|
| Structure | Anything that links to nothing above it | Show me every priority serving no objective |
| KPI cards | Which KPIs are missing a baseline or a source | List the incomplete cards |
| Correlations | Primaries with no mechanism or kill signal | Show me the correlations that would fail the gate |
| People | Rows with no owner, and people with too many | Show me everyone owning more than two rows |
| Agreement | Catchball sessions not closed by agreement | Show me which negotiations are still open |
| Operation | Rows with no phased plan values or no cadence | List the rows that are still an annual number |
| Lock itself | A state, where targets need a recorded reason to move | Lock it, then try to change a target |
| Partial lock | Lock most of it and carry named open items with dates | Lock with one escalation outstanding |
Row 7 is what separates a locked plan from a plan everyone agreed to stop editing.
Row 8 is the practical one. Almost every planning season ends with one or two items genuinely unresolved, and a tool that only offers all-or-nothing lock pushes teams into locking something untrue or not locking at all. Both end the same way: January spent finishing the plan.
What locking too early costs
The worked example found three fails in week eleven and fixed them inside the week. Had the gate not run, those three arrive in the January review instead: two KPIs that cannot be reported, a priority whose owner is “commercial” rather than a person, and a target still being argued. The review then spends its first hour on plan repair rather than performance, and does it again in February.
That is the quiet cost of a plan that was agreed but not ready: the first quarter is spent finishing it, which is the same quarter the plan was supposed to be moving. Hoshin planning mistakes that show up in March traces the rest of what surfaces from a weak gate.
On licensing, the models differ in shape rather than scale. OnStrategy lists $200 per month for unlimited users. i-nexus publishes a free Starter tier for up to five users and $28 per user per month at its Standard tier. The gate reaches every row and every owner, including the sites in the roll-up check, so a per-seat model prices the completeness the check depends on.
Profit.co publishes two plans, Essentials and Enterprise, quoting against modules and scale rather than a public seat rate. Hoshin Kanri sits in Essentials with OKRs and Goals, Balanced Scorecard and Strategy Roadmap, so the locked plan is the plan the teams run their quarters on rather than a document they report against. What drives the cost of hoshin kanri software covers the rest to total before comparing quotes.
The 30-day free trial is long enough to run the thirteen checks against a current matrix and see how many it would fail today.
Locking in Profit.co
In Profit.co the locked plan is the operating plan: the same objectives, priorities, KPIs and graded correlations on the X-matrix, with monthly plan values in the bowler chart, owners named on every row and check-ins and huddles running from the start of the year.
Catchball closing is explicit: a session closes when every participant has agreed, so the gate’s “catchball closed” check reads straight from the sessions. After lock, target changes go back through catchball with the original kept as baseline, and correlation changes are written to each correlation’s strength history. More on the planning software page.
That is the table above expressed as behavior rather than policy. Countermeasures and team key results keep moving; a target does not change without a route and a reason. Which is also why next October’s reflection has something to read, and why replanning mid-year stays a decision rather than a quiet edit.
Free Product Demo
Run the lock gate on your plan
A 30-minute walkthrough, tailored to your team.
Book a free demo
What happens in the call
How your teams set goals, track progress and review results.
Strategy, performance, projects and engagement on one platform, with AI.
A practical path to align teams, execute faster and drive results.
Questions this article answers
That objectives, targets, correlations and owners now change only through the quarterly review with a recorded reason. Countermeasures, tasks and team key results still change freely.
Links from priorities to objectives to breakthroughs, complete KPI cards, gated primary correlations, one owner per row, resolved capacity, closed catchball, phased plan values and a set check-in cadence.
Yes. Lock everything ready and list open items, such as a target awaiting a budget decision, with dates, rather than holding the whole plan.
Week 11 of a twelve-week calendar, after cascade and before deployment, so that failures can be resolved or contained inside the same week.
Agreement means everyone accepts the content. Readiness means the first monthly review could actually run from it: baselines exist, every row has one owner, plan values are phased and check-ins are dated.
Answer each gate check as a query rather than a manual audit, and make lock a real state: targets and grades change only with a recorded reason, while countermeasures and team key results stay free.
