Implementation plan

An implementation plan turns a commitment — a countermeasure, a kaizen charter's rollout, a Hoshin catchball agreement — into a real, checkable sequence of milestones converging on a target completion date. Each milestone names who it depends on: your own team, an outside party, or both jointly. Days remaining, which milestones are already overdue, and which have no named owner yet are all computed from what you enter, not asserted. The one check unique to this tool: if every milestone belongs to your own side, that's flagged directly — a plan with zero cross-functional dependency usually means nobody outside the initiating team has actually committed to anything yet, not that the plan is simpler than most.

Milestones

Describe the commitment roughly — a first-pass milestone sequence gets drafted, each with who it depends on (you, them, or joint) and a target date. Assign a real named owner to each afterward; that's never guessed for you.

Coverage

Computed from your milestones and dates — nothing to enter here.

Risks

Drafted alongside the milestones — what could realistically cause this plan to slip.

Narrative

Drafted from the plan actually built.

The example below rolls out a new andon escalation rule across a fabrication cell's three shifts — one milestone is already honestly overdue, one has no owner named yet, and the plan is genuinely cross-functional rather than something the improvement team alone could execute.

• Writing every milestone as something your own team does.

  A plan where every milestone belongs to your own side isn't actually testing whether anyone else has really committed — that's exactly what the mutuality check exists to catch, not a sign the plan is simpler than most.

• Leaving a milestone unowned and moving on rather than naming the gap.

  An unowned milestone isn't a detail to fill in later — it's a real risk to the completion date, and this tool's whole point is surfacing that honestly instead of letting a plan look more complete than it is.

• Writing a risk as vague as 'things could change.'

  A risk that isn't specific enough to actually watch for isn't useful — name the real thing that could slip this plan, the same discipline this catalogue applies to every other finding.