Hoshin Kanri — X-Matrix

Hoshin Kanri ("direction management") is how an organisation turns a long-term direction into this year's work, without the connection between the two getting lost. The X-Matrix is its signature artifact: one page linking long-term breakthrough objectives, this year's annual objectives, the improvement strategies that will move them, and the metrics that prove it — with a correlation grid at each intersection showing exactly how strong each connection actually is. Catchball is the negotiation that gets you there: objectives don't arrive fully formed from the top, they're proposed, pushed back on, and refined at every level before they're locked in.

True North

One or two sentences: why does this plan exist? Every objective on the page should trace back to this. If you can't draw that line for an objective, it doesn't belong on this plan.

Long-term breakthrough objectives

Where does the organisation need to be in three to five years? Give each one a target that's unambiguous when reached — "be more competitive" isn't checkable; "grow market share from 12% to 20%" is.

Annual objectives

What has to be true by the end of this year to be on track for the long-term objectives? These are what actually get worked this year — the long-term objectives set direction, the annual objectives set the pace.

Catchball

The record of who proposed each annual objective, how another level pushed back, and what was actually agreed. This isn't paperwork — an objective that was never negotiated with the people who have to deliver it is a wish, not a commitment.

Strategies

The improvement initiatives that will actually move the annual objectives. Name an owner for each — a strategy with no owner doesn't happen, it just sits on the page looking like progress.

Metrics

What you'll track to know the strategies are working, each with a target. If a strategy has no metric watching it, you'll find out it failed only when the year is over.

X-Matrix correlations

The heart of the X-Matrix: which long-term objectives each annual objective actually supports, which metrics track each objective, which strategies move each metric, and which strategies support which long-term objectives. Mark primary, secondary, or weak — and leave a pair blank when there's genuinely no connection. A matrix where everything correlates with everything isn't rigorous, it's decoration.

The example below is a full X-Matrix: long-term objectives and annual objectives with real targets, a catchball record showing the negotiation behind them, strategies with named owners, metrics with targets, and a correlation grid that's honest about which connections are strong and which are weak — not a fully-shaded matrix pretending everything connects to everything.

• Objectives read like a mission statement, not a target.

  "Be a leader in customer experience" can't be checked. If two people could reasonably disagree about whether it was achieved, it isn't specific enough to plan against yet.

• The correlation matrix is fully shaded — everything connects to everything.

  A matrix that looks that dense usually means the correlations weren't actually thought through. The useful information in an X-Matrix is which connections are weak or absent, not just which are present.

• Catchball is skipped and the matrix goes straight from an executive offsite to a finished document.

  The people asked to deliver the strategies never got a chance to push back on feasibility. The plan looks complete and isn't tested.

• A strategy has no owner, or a metric has no target.

  This is exactly what the specificity check exists to catch — an unowned strategy or untargeted metric is a line on a page, not a commitment.

• One person builds the whole X-Matrix alone to save time.

  Hoshin's value comes from the organisation aligning around the plan, not from the document itself. A matrix nobody else was involved in building has no more authority than an opinion.