Hoshin Kanri for COOs: close the gap between strategy and Linear
The COO owns the gap between what the executive committee decided and what the teams actually ship. Hoshin Kanri closes it structurally: every strategic initiative in the X-Matrix carries a link (external_ref) to its execution project in Linear, Jira or Asana, and a weekly agent sync reconciles the two — progress pulled automatically, drift flagged, review agenda drafted. No re-typing, no parallel reporting.
This page describes the setup for an operations leader: what lives where, what the weekly loop looks like, and what it replaces.
The two-systems problem
Strategy lives in slides; execution lives in Linear or Jira. Between the two sits a translation layer — usually you, the COO, rebuilding a status deck every month from sprint boards, chat threads and memory. The symptoms are always the same:
- Projects in the execution tool that nobody can trace to a strategic objective — shadow work that consumes real capacity.
- Strategic initiatives with no execution project behind them — strategy theater that looks green until the quarterly review.
- Status meetings that spend their time collecting status instead of arbitrating, because the data was never fresh in one place.
Link, don't duplicate: external_ref
Hoshin Kanri draws a hard boundary. The X-Matrix holds the why and the what: objectives, initiatives, KPI targets, correlations, reviews, decisions. Your execution tool keeps the how: issues, sprints, assignments, daily tracking. The bridge is one field — each initiative carries an external_ref pointing to its execution project.
There is deliberately no task management, no Gantt chart and no duplication of issues in the matrix. Anything that would require your teams to maintain the same information in two tools is a design error the platform refuses to make — because that duplication is exactly what kills strategy tools by March.
The weekly loop, run by an agent
Connect Claude (or Cursor) to the Hoshin Kanri MCP server next to your Linear or Jira MCP — OAuth, about two minutes each. Then the built-in weekly_sync prompt runs the loop every Monday:
- Reads the initiatives in the South quadrant and their external_ref links.
- Pulls progress and blockers from each linked project through the execution tool's own MCP.
- Updates initiative statuses in the matrix — the strategy is now as fresh as the sprint board.
- Flags drift in both directions: execution projects mapping to no initiative, initiatives linked to nothing.
- Drafts the weekly review agenda: off-track KPIs, blocked initiatives, decisions pending.
KPIs that prove something
The West quadrant holds the KPIs, and each one is correlated to the objectives and initiatives it is supposed to prove. That correlation is what turns a dashboard tour into a review: the question is never "what does this number say?" but "which strategic bet does this number confirm or refute?".
The completeness score audits the indicator set itself — including the leading/lagging balance. An operations dashboard that only contains lagging indicators tells you where you failed after the fact; the score pushes you to instrument the leading signals while there is still time to act.
What this replaces — and what it does not
Replaced: the Sunday-evening status deck, the meeting that collects status instead of deciding, the quarterly surprise where an initiative turns out to have been dead for two months, and the role of human middleware between strategy and execution.
Not replaced: your project management tool (Linear, Jira, Asana keep the how), your BI stack (measurement data and time series stay where they are). Hoshin Kanri holds the strategy layer and stays linked to both.
Frequently asked questions
Does this replace Linear, Jira or Asana?
No — by design. Your execution tool keeps tasks, sprints and issues. Hoshin Kanri keeps objectives, initiatives, KPIs and reviews, and each initiative links to its execution project via external_ref. The weekly_sync agent prompt keeps the two aligned without double data entry.
What is external_ref exactly?
A field on each matrix initiative that stores the reference of its execution project (a Linear project, a Jira epic…). It is exposed everywhere — editor UI, API, exports, MCP tools — so both humans and agents can jump from a strategic initiative to the work behind it.
What happens when a team works on something outside the strategy?
The weekly sync surfaces it: an execution project that maps to no initiative is flagged as potential strategy drift. Then it is a management decision, made in review with the data on the table — link it to an existing initiative, make it a new one, or stop it.
Which execution tools can the agent read?
Any tool that exposes an MCP server — Linear and Jira are the common cases. MCP is an open standard, so the integration is not bespoke per tool: the agent reads your matrix through the Hoshin Kanri MCP and the execution data through each tool's own MCP.
How much time does the weekly loop actually take?
The agent prepares everything — statuses updated, drift flagged, agenda drafted. The human part is the review itself: arbitrating the exceptions, typically 30 minutes. The point is that operations leadership spends its time deciding, not collecting.