Policy Deployment From Board to Shop Floor

Most businesses have a strategy of some kind. Far fewer have a way of making sure it changes what people do on a Tuesday afternoon. Policy deployment, often called hoshin kanri, is a simple discipline for closing that gap. It takes a 1, 3 and 5-year plan, reduces it to a handful of objectives, gives each one a measure and a named owner at each level of the business, and reviews progress regularly. The difficulty is keeping it simple and keeping it going.

Start with the long view, then narrow it

Policy deployment starts at the top. The board agrees where the business needs to be in five years: what it sells, to whom, at what scale and with what capability. That is then broken into what must be true in three years and what must be achieved this year.

The one-year plan is where most of the effort goes. From everything the business could do this year, the board picks a small number of breakthrough objectives. These are the things that, if done, move the business noticeably towards the three-year position. Everything else is either business as usual or deliberately put on hold.

A typical set for a manufacturer might include:

• Reduce dependence on the largest customer by winning new work in a target sector.
• Bring on-time delivery up to the level customers now expect.
• Achieve a new certification needed to bid for that work.
• Develop a second tier of managers who can run the business without the owner.

Four is enough.

Cascade objectives, measures and owners

Each board-level objective is then broken down. Operations takes the delivery objective and asks what it needs from scheduling, the toolroom, maintenance and purchasing. Each of those areas turns it into something they can act on and measure. The cascade continues until it reaches the teams and cells on the shop floor.

At every level, each objective should have three things:

• A measure that shows whether it is moving, with a target and a date.
• A named owner, a person rather than a department.
• A short list of actions that the owner is accountable for delivering.

The cascade works best as a conversation in both directions. Senior managers propose; the people below test whether the targets are realistic and what they need to deliver them; the plan is adjusted. Hoshin kanri calls this "catchball". In plain terms it means the people who have to do the work get a say in the plan before it is fixed. That is what turns a target handed down into one people own.

Our article on organisational design and accountability covers the structure that sits underneath this. Policy deployment depends on clear roles; it cannot fix a business where nobody knows who decides what.

Review regularly, at every level

A plan without review is a document. Policy deployment relies on a fixed rhythm of review, with each level looking at its own measures:

• Board: monthly, looking at the breakthrough objectives and whether each is on track.
• Departments: weekly, looking at their own measures and the actions behind them.
• Shop floor: daily, at a board in the area, looking at the few measures that team can influence that day.

The review should be short and should always ask the same questions. Is the measure on track? If not, why not, and what are we doing about it? Who needs help? A measure that stays off track needs proper problem solving.

Where it usually goes wrong

In our experience, policy deployment fails in a few predictable ways.

• Too many priorities. A board that cannot choose ends up with fifteen objectives, each owned by someone already at full stretch. Nothing gets the attention it needs. Choosing what not to do is the hardest and most important part of the exercise.
• Objectives with no owner. An objective assigned to "the management team" or "quality" belongs to no one. If a name cannot be put against it, it will not happen.
• Plans that never reach the shop floor. The board has a plan, the managers have a version of it, and the operators have never seen it. When the people doing the work cannot say how their measures connect to the business's goals, the cascade has stopped somewhere above them.
• Measures that cannot be acted on. A cell team cannot influence group profit. They can influence scrap, set-up time and first-off approval. Each level needs measures it can actually move.
• Review that lapses. The first few months go well, then a busy period arrives, reviews get cancelled and the plan quietly dies. The rhythm has to survive the busy months, because that is when it is most useful.

Keep it on one page

The best test of a deployed plan is whether each level can show its part on a single page: objectives, measures, owners, actions and current status. If it needs a thick binder, it is too complicated to be used. Our operational excellence self-assessment is a useful way to decide where to focus first.

Policy deployment is part of our governance and organisational design work, from setting the 1, 3 and 5-year plan through to building the review rhythm. If that would help your business, get in touch.