Last Planner System Software Quick Reference

The Last Planner System is not complicated. What trips people up is the day-to-day muscle memory: which meeting comes first, what a promise actually has to satisfy before you make it, how to score the week honestly, and what to do when the number comes back ugly. This is the reference I wish someone had taped to the inside of my trailer door when I first ran a pull plan. Keep it handy, use it in your weekly meeting, and hand it to the new engineer who's about to facilitate their first session and looks terrified.

The idea in one breath

Traditional CPM scheduling tells you what should happen. Last Planner is about what can and will happen — planned by the people who actually do the work. You plan in more detail as the work gets closer, you clear the roadblocks before you commit, you make promises you can keep, and you learn from the ones you break. That's the whole thing. Everything below is just the mechanics that make it stick.

The planning levels, and how far out each one lives

Think of it as one schedule seen at four zoom levels. Confusing them is the most common rookie mistake.

• Master / milestone schedule — the whole job. Set once, revisited at milestones. This is your contract skeleton, not your working tool.
• Phase (pull) plan — a chunk of the job (say, "podium deck to weather-tight"), planned backward from a milestone with all the trades in the room. This is where handoffs get negotiated.
• Look-ahead — the make-ready window, typically 6 weeks out, sometimes 3 or 4 depending on how fast your job moves. This is where you find and kill constraints. If your look-ahead is only a to-do list of what's coming, you're using it wrong.
• Weekly work plan — the commitments for the next 5–7 days. Only work that is genuinely ready shows up here. This is the promise sheet.

Rule of thumb on the look-ahead window: pick a horizon that's a little longer than your slowest lead time for the constraints you actually hit. If a long-lead material or a plan-check turnaround runs five weeks, a three-week look-ahead can't protect you — you'll find the constraint too late to clear it. When in doubt, go six weeks.

The constraint checklist — run every activity through it

Make-ready is the heart of the system, and it lives or dies on whether you actually screen every incoming activity against the same categories. Memorize these seven. When someone in the look-ahead meeting says "yeah, we're good for that," walk it down the list out loud:

• Prerequisite work — is the preceding activity truly finished and inspected, not "basically done"?
• Information — approved submittals, RFI answers, final dimensions, a signed-off layout. Missing an ASI is a constraint even if the crew is standing there.
• Materials — ordered, confirmed ship date, and staged where the crew needs them. "On order" is not "ready."
• Labor — the right crew, the right size, actually available that week and not pulled to another job.
• Equipment — lift, scissor, welder, crane pick window. Shared equipment is a constraint hiding in plain sight.
• Space / access — is the area physically clear, and does another trade have it that week? Two crews fighting over the same room is a sequencing failure, not a personality problem.
• External — permits, inspections, utility tie-ins, weather. These have the longest lead times and the least flexibility, so screen them earliest.

Every constraint gets a name and a date — who clears it and when. A constraint owned by "the team" is owned by nobody. The single biggest lift you get from good look-ahead scheduling software like LookAheadWall is that a constraint stops being a note in someone's phone and becomes a visible, assigned, dated item you can't lose track of between meetings.

What makes a promise "reliable"

The word commitment gets thrown around loosely on jobsites. In Last Planner it has teeth. Before an activity earns a spot on the weekly work plan, the promise behind it should pass four tests:

• Definition — everyone means the same thing. "Rough-in level 3" is vague; "rough-in electrical, north wing, rooms 301–318, ready for cover inspection" is a promise.
• Soundness — the prerequisites are clear, the sequence is right, and all seven constraints above are removed. No open constraints, no promise.
• Size — it fits inside the week. If it can't finish in the commitment period, break it down until it can. Chunks you can actually complete are what make PPC meaningful.
• Learning — the outcome is checkable. You'll know Friday whether it happened, so you can learn from it either way.

Here's the cultural part that no software solves for you: reliable promising only works if it's safe to say "no, not next week." The first time a foreman gets burned for honestly declining a commitment they weren't ready for, you've taught the whole table to over-promise. Protect the people who tell you the truth about what isn't ready.

PPC — measure it, and read it right

Percent Plan Complete is the pulse of the system:

PPC = (commitments completed as promised ÷ total commitments made) × 100

Two things people get wrong. First, it's all-or-nothing — 90% done is a zero. That feels harsh until you realize the point isn't to grade effort, it's to expose planning reliability. Partial credit hides the misses you need to see. Second, a low PPC is not a stick to beat crews with. It's a diagnostic. If you turn PPC into a report card that gets people yelled at, they'll start committing to less and less so the number looks good — and you've killed the whole loop.

For a healthy job, PPC in the high 70s to mid 80s is a realistic, honest range. A team parked at a suspiciously smooth 95% every single week usually isn't planning brilliantly — they're sandbagging their commitments. And a chart that only ever goes up is worth a second look; real jobs have bad weeks. The trend matters more than any one number.

Variance analysis — the part everyone skips

PPC tells you that you missed. Variance analysis tells you why, and it's where the actual improvement lives. Every incomplete commitment gets a reason, and you tally the reasons over time. The categories mirror the constraint list — prerequisite work, information, materials, labor, equipment, space, external — plus the honest ones like "we changed our mind" or "underestimated the duration."

Do this for a month and a pattern will jump out. Maybe 40% of your misses trace back to information — late submittals, unanswered RFIs. That tells you the fix isn't in the field at all; it's upstream in the submittal log. Chasing crews harder would have done nothing. When your look-ahead data is captured cleanly, that root-cause pattern is right there to read instead of buried in a stack of dry-erase photos. Don't let the meeting end without at least one countermeasure aimed at the top recurring reason. One real fix a week beats ten good intentions.

Running the weekly meeting — a facilitation script

Keep it to 45 minutes. The trades will forgive a lot if you respect their time and run a tight room. A sequence that works:

1. Score last week first (5 min). Read the PPC, and for every miss, capture the reason on the spot. Do this before anything else so nobody quietly forgets what they promised.
2. Walk the look-ahead (15 min). Go activity by activity through the make-ready window. Every open constraint gets an owner and a due date. This is 90% of the value of the meeting.
3. Build next week's work plan (15 min). Only activities with zero open constraints get committed. If a trade wants to commit to something still constrained, that's your cue to say "let's clear X first."
4. One improvement (5 min). Pick the top recurring variance reason and name one action to attack it. Assign it.
5. Confirm and close (5 min). Read back the commitments so everyone hears their own promises out loud. Verbal readback is where soft promises get exposed.

Facilitator's job is to ask questions, not to fill in the schedule for people. The moment you start making commitments for a sub, you own their misses and they own nothing. Put the marker in their hand.

Common failure modes, and the tells

• The look-ahead is just a countdown, not a make-ready tool. Tell: constraints get "discovered" in the weekly meeting instead of cleared a week earlier. Fix: screen the whole window every session.
• Suspiciously high PPC. Tell: the number never dips and the schedule never slips faster than the plan says. People are committing to less than they could. Push for honest, fuller commitments.
• Constraints with no owner. Tell: the same item shows up "in progress" three weeks running. Every constraint needs a name and a date, full stop.
• Variance reasons never get acted on. Tell: you're logging why you miss but the top reason is identical month over month. Data without a countermeasure is just paperwork.
• Superintendent runs the whole plan. Tell: the foremen sit quietly and nod. If the last planners aren't planning, it isn't Last Planner.

Where the software actually earns its keep

You can run all of this with sticky notes and a whiteboard — plenty of great teams started exactly there, and there's no shame in it. What a purpose-built tool buys you is that nothing falls through the cracks between Fridays: constraints stay assigned and visible, the weekly work plan is location-based so you can see two trades stacking in the same area before it becomes a fight, PPC and variance reasons roll up automatically instead of dying on a whiteboard, and the plan reaches every foreman's phone instead of only the folks who made it to the trailer. That's the niche LookAheadWall is built for — turning the look-ahead and weekly work plan into something the crew can see and act on, with the mobile companion putting the same plan in the crew leader's pocket. But the discipline comes first. The best software in the world won't fix a team that doesn't screen constraints or won't tell the truth about a miss. Get the habits right, and the tool just makes them faster.

Print this, keep it close, and run your next weekly meeting off it. The system rewards teams that keep showing up and doing the boring parts — screening every constraint, scoring every miss, fixing one root cause at a time. Do that for a few months and the reliability shows up in the field, where it counts.