Hoshin Kanri · Policy Management
Hoshin Kanri / Policy Deployment
What is hoshin kanri?
- Limit annual key policies to 3–5
- Align top and bottom through repeated catchball
- Check alignment on a single X-matrix
Putting hoshin kanri in place changes the nature of the opening meeting of the year, not the thickness of the plan. The session that used to add up each department's initiatives becomes the session that decides which initiatives you will not pursue this year. Once three or four breakthrough goals are fixed, the dozens of others automatically become 'things we won't do this year,' and the real effect is that this list is written down rather than agreed verbally. Conflicts over resource allocation move up from year-end performance reviews to the point when policy is set at the start of the year.
In practice, work is split into two tracks: breakthrough initiatives and daily management. The norm is no more than 3–5 breakthrough initiatives company-wide and 2–3 per division. Metrics that are already managed, such as yield and on-time delivery, move to daily management and are handled through standard work and exception reporting. Catchball typically runs 2–3 rounds over 6–8 weeks. Lower-level units don't just reply, 'we received the target.' They come back with evidence on which conditions would make it work and which would not. Correlation strength marks such as ◎, ○, and △ go at the X-matrix intersections, so initiatives without metrics and metrics without owners show up at a glance.
It is often explained alongside OKR, but the operating conditions differ. OKR resets every quarter, leaves goal-setting authority with teams, and as a rule is kept separate from performance reviews, whereas Hoshin Kanri is fixed annually and tied directly to budget and capital-investment decisions. At year-end, only the missed items are pulled together for a reflection (hansei) meeting, and the causes are fed into next year's policy as inputs. That is why Hoshin Kanri fits manufacturing and engineering organizations whose investment decisions are locked into annual cycles, while OKR fits organizations whose product specs shift every quarter.
The most common failure is catchball degenerating into a game of splitting up numbers. When leadership throws out “20% company-wide growth,” each division simply cuts it into 20% slices and hands them to teams. That isn’t deployment; it’s copying. If you cut the number of initiatives without building a daily management system, the dropped work comes back three months later as “urgent initiatives,” and the list swells back to twenty. An X-matrix drawn once at the start of the year and pinned to the wall becomes decoration if no one opens it at quarterly reviews. And if analysis of missed targets turns into blame, a downward distortion sets in the following year, where only targets certain to be met get submitted.
John KimAuthor of B2B Marketing Design and CEO of Nanum Consulting
Ilyoung JungExpert in smart factories, equipment management, and the Fourth Industrial RevolutionIt took shape in the 1960s during total quality control (TQC) at Japanese companies and spread through firms such as Bridgestone and Toyota. In English it is usually called Policy Deployment.
Let's walk through a hypothetical case. Company A is a tier-2 auto parts supplier with 320 employees and KRW 110 billion (≈ US$82,298,400) in annual revenue. Every February it put 27 strategic initiatives into its management plan, yet only 6 were completed by year-end, and its division heads called them 'not projects but homework.' When the CEO ordered a four-month shift to hoshin kanri, the planning office rebuilt the process from scratch.
- The head of planning audited all 27 initiatives and kept only those that directly affected revenue and profit, which left 4. Of the remaining 23, 16 were routine work already running as department KPIs, so they were dropped from the policy list entirely.
- In the first round of catchball, the head of production put his foot down: “We can’t reach 92% on-time delivery without adding two machines.” Management lowered the target to 89% and, in exchange, approved KRW 800 million (≈ US$598,500) for the added equipment in March.
- Over two more rounds of back-and-forth at three-week intervals, the sales division asked to cut its new-account target from 12 to 7 in exchange for authority over per-item price negotiation, and the CEO agreed on the spot.
- Once the 3-year goals, the 4 breakthrough initiatives for this year, 9 metrics, and the executive owners were all on a single X-matrix, a cost-reduction initiative that had a metric but no owner came to light, and the purchasing manager took it on that same day.
- Monthly review meetings were capped at 40 minutes, and teams reported only where they were stuck instead of progress percentages. When delays at the mold subcontractor came up three months in a row, the CEO approved dual-sourcing right in the meeting.
- At year-end, 3 of the 4 initiatives had beaten their targets. On-time delivery rose from 81% to 90%, the initiative completion rate climbed from 22% to 75%, and only 2 'urgent initiatives' interrupted the year.