The Last Planner System: An Overview for Schedulers

The Last Planner System is a way of planning construction work that puts the people closest to the work in charge of the short-term plan, and measures whether they do what they said they'd do. If you're a scheduler, it can look like a rival to CPM. It isn't. It's a set of planning layers that sit under the master schedule and deal with what CPM handles badly: whether the work can actually be done next week. This post explains the pieces, how they connect, and where your schedule fits.

The core idea

Traditional scheduling tells crews what they should do, based on the logic and the dates. Whether the work can be done depends on things the schedule doesn't track well: whether the submittal came back, whether the material is on site, whether the previous trade left the area clean, whether the inspector is booked. Crews then do what they will do, which is some subset of what can be done. The Last Planner System tries to close the gap between should, can, will, and did.

The "last planner" is the person who makes the final assignment of work to a crew: usually a foreman or trade superintendent. The system's argument is that plans made by that person, with constraints cleared in advance, are more reliable than plans handed down from the office.

The Lean Construction Institute developed and promotes the system, and it's usually what U.S. teams mean when they talk about pull planning and weekly work plans. "Last Planner" is a registered trademark of LCI.

The five levels

The system is usually described as five connected planning conversations, from long-range to daily.

Some descriptions add daily huddles as a separate level. The structure is the same either way.

Master scheduling

This is the schedule you already know: the CPM network, contract milestones, and critical path. In a Last Planner job its role is to set the milestones each phase will plan toward and to keep track of the whole-job logic. It doesn't need to carry crew-level detail, and usually shouldn't. See schedule level of detail for how far to go.

Phase scheduling

For each major phase (foundations, structure, building enclosure, interiors by floor), the trades meet and pull-plan backward from a milestone. Each handoff is a request from the downstream trade and a promise from the upstream one. The result is a sequence the trades built themselves, plus a first list of constraints. How to run a pull planning session covers the mechanics.

Lookahead planning and make-ready

Each week, the team looks several weeks ahead, breaks phase tasks into operations, and screens each one: is anything stopping it from starting on time? Constraints get owners and due dates. The aim is that by the time a task enters the weekly plan, it's ready to go. This step is where most of the value lives, and where most teams cut corners. Make-ready planning and constraint logs goes into it.

Weekly work planning

Each foreman commits to specific tasks for the coming week, drawn only from the ready work in the lookahead. A commitment should meet a few quality tests:

• Definition. It's specific enough that everyone can tell when it's done ("hang board rooms 301-306", not "work on L3").
• Soundness. Its constraints are cleared.
• Sequence. It's the right next work, in the right order.
• Size. It fits the crew's capacity for the week.

The foreman, not the GC, decides what to commit to. If a foreman won't commit to something, that's information: either a constraint is hiding, or the plan is unrealistic.

Learning

At the end of the week, each commitment is marked done or not done. The share completed is Percent Plan Complete (PPC). For each miss, the team records a reason for variance and, for the recurring ones, digs into root cause. Over time the reason categories show where the system leaks: maybe it's submittals, maybe it's prerequisite work, maybe it's manpower. Percent Plan Complete and reasons for variance covers how to measure and use it.

A short example

A drywall foreman on a four-story office building, week 14:

Lookahead (weeks 14-17) for L3 board, screened:
  Hang board rooms 301-306     constraints: in-wall inspection   status: cleared
  Hang board rooms 307-312     constraints: in-wall inspection   status: scheduled week 15
  Tape and finish 301-306      constraints: hang complete        status: depends on week 14
  Hang board 313-324           constraints: plumbing rough-in    status: plumber says week 16

Weekly commitment for week 14:
  1. Hang board rooms 301-306
  2. Stock board on L3 east for 307-312

End of week 14:
  1. Done
  2. Not done -- hoist booked by another trade Thursday and Friday
  PPC for drywall this week: 1 of 2 = 50%
  Reason: shared equipment / logistics

The foreman didn't commit to 307-312 because the inspection wasn't cleared. That's the system working. The stocking miss is a real failure, and the reason points at a coordination fix: hoist time needs to be on the weekly plan.

Where CPM and Last Planner disagree, and where they don't

Schedulers sometimes hear Last Planner described as anti-CPM. In practice the two answer different questions.

• CPM answers: given the logic and durations, when will the project finish, and which work drives that date? It's good at the whole-job view and at quantifying delay.
• Last Planner answers: what work can we reliably do next week, and why didn't last week's plan happen? It's good at short-term reliability and at surfacing constraints.

The friction is real but manageable:

• Durations. Pulled durations from foremen are often different from the CPM durations. That's useful feedback for the master schedule.
• Logic. Phase plans often reveal sequence the CPM got wrong or oversimplified. Update the CPM logic, not just the dates.
• Float. Trades planning a phase don't usually see total float. Give it to them, at least for critical and near-critical work, so they know where a slip hurts the whole job.
• Two sources of truth. If the pull plans and the CPM schedule drift apart, people stop trusting both. The fix is a regular reconciliation, covered in connecting pull plans back to the master schedule.

What it takes to work

The system is mostly habits, not software. The habits that matter:

• Weekly planning meetings that happen every week, at the same time, with foremen present.
• Commitments made by the people who'll do the work, from screened work only.
• An honest PPC, including the weeks it's bad.
• Reasons for variance recorded every week and acted on.
• A constraint log someone actually maintains.
• A master schedule that gets updated with what the phase plans learn.

Tools help with the bookkeeping: tracking commitments, computing PPC, keeping the constraint log in one place, tying tasks back to CPM activities. schedule.construction has a pull planning board and weekly work plans with PPC that sit on the same data as the CPM schedule, which is one way to keep the two from drifting. A whiteboard and a spreadsheet will also work if someone owns them.

Key takeaways

• Last Planner is five planning levels: master, phase, lookahead, weekly, and learning. CPM is the master level.
• The goal is reliable short-term plans: only commit to work whose constraints are cleared.
• The last planner is the foreman who assigns the crew, and that person makes the weekly commitment.
• PPC and reasons for variance show where the planning system leaks, and point at what to fix.
• Keep the CPM schedule and the phase plans reconciled, or you end up with two schedules nobody believes.