The Last Planner System gets talked about like a religion and used like a rumor. Half the superintendents I meet can quote the four levels, and about a fifth of them actually run it on a jobsite. The gap between knowing it and doing it is almost always in the follow-through — the make-ready work, the honest weekly commitment, the willingness to write down why last week fell apart. Software doesn't close that gap by itself. But the right tool takes the bookkeeping off your plate so you can spend your attention on the part that matters: getting real commitments from the people who actually swing the hammers.
This is a working superintendent's take on what the Last Planner System really asks of you, where crews trip over it, and what a scheduling tool needs to do to be worth the login.
What Last Planner Actually Is (and Isn't)
The core idea is embarrassingly simple: the people who do the work should plan the work. The "last planner" is the foreman or crew lead who makes the final assignment — not the PM in the trailer, not the master schedule in the corner office. Everything in the system exists to move planning down to that person and make their commitments reliable.
It runs on four nested levels of detail, each one filtering into the next:
- Master schedule — the milestone-level should-do. Contract dates, phase handoffs. You built this at buyout and you touch it rarely.
- Phase pull plan — the team works backward from a milestone (say, "conditioned air on floor 3") and hangs sticky notes for each handoff until the sequence and durations are agreed by the trades themselves.
- Lookahead (make-ready) — a rolling 6, 4, or 3 week window where you screen every upcoming activity for constraints and knock them down before the work is due.
- Weekly work plan — only the activities that are genuinely ready. This is what crews commit to on Monday.
What it isn't: a Gantt chart with better marketing. A traditional CPM schedule tells you what should happen based on logic and float. Last Planner is about what can and will happen based on reality — materials on site, prior trade complete, area released, inspection passed. The two live side by side. The master schedule sets the target; Last Planner is how you hit it week to week without lying to yourself.
Pull Planning: Where the Sequence Gets Real
The pull plan is the most valuable meeting you'll run all year, and most people rush it. You put the milestone on the right side of the wall, hand every trade foreman a stack of cards, and work backward. Each trade writes what they need done before they can start, and how long their piece takes. The drywall foreman says he needs the wall signed off by inspection. The inspector's sign-off needs rough-in complete. Rough-in needs the layout. And so on, right to left, until the whole handoff chain is on the wall in the trades' own words.
Two things happen in a good pull plan that never happen in an office-built schedule. First, the durations are honest, because the person committing to them is standing right there and their peers are watching. Second, the conflicts surface immediately — the electrician and the plumber both need the same ceiling space in the same week, and now you find out in a conference room instead of on the deck.
A rule of thumb: build in a handoff buffer between trades, not just activity durations. Frame-to-rough-in usually wants a day or two of slack for cleanup, punch, and inspection. Overhead MEP before ceiling grid wants a real gap so the last hanger isn't fighting the grid crew. When you compress those buffers to zero to make the master schedule look good, you're not scheduling — you're gambling.
Make-Ready: The Part Everyone Skips
If Last Planner has a beating heart, it's make-ready planning in the lookahead window. This is where you take every activity coming up in the next three to six weeks and ask one question of each: what would stop this from happening? Those blockers are your constraints, and the classic categories are worth memorizing because crews hit the same seven over and over:
- Prior work — is the predecessor actually complete, not "basically done"?
- Materials — on site, or a submittal-and-lead-time gap you didn't catch?
- Information — RFI answered, detail released, latest drawing in hand?
- Equipment — lift, crane time, hoist window reserved?
- Labor — right headcount and the right certifications?
- Space — area physically released and not stacked with another trade?
- External — permit, inspection, owner decision, utility.
The discipline is that a constraint has an owner and a promised clear-by date, and you screen it in the lookahead every single week until it's gone. An activity does not graduate to the weekly work plan until its constraints are clear. That's the whole trick. The reason PPC (we'll get there) collapses on most jobs is that people put work on the weekly plan that was never actually ready — the pipe hadn't landed, the inspection wasn't scheduled — and then act surprised when it doesn't happen.
Make-ready is unglamorous. It's phone calls, submittal chasing, and nagging. But an hour spent clearing constraints three weeks out saves a crew standing around on Thursday with nothing to do. That's the trade you're making.
The Weekly Work Plan and Reliable Promising
By the time you get to the weekly work plan, the hard thinking is done. This plan contains only ready work, and the crews commit to it out loud. That commitment — "yes, my guys will have that wall closed by Wednesday" — is what Last Planner calls a reliable promise, and it only means something if it's freely given and can be declined.
This is where superintendents new to the system get it wrong: they turn the weekly meeting into a directive. You will finish this. That's not a promise, it's an order, and an order the foreman didn't buy into is exactly the commitment that fails. The move is to ask, listen, and let the foreman say "no, I can't, because X" — because now X is a constraint you can go clear. A "no" on Monday is a gift. A false "yes" is a Thursday problem you don't see coming.
Keep the weekly plan at the level the crew works: by area and by trade, in day-sized chunks. If an activity is bigger than a week, break it. A weekly work plan full of two-week bars isn't a commitment tool, it's just the master schedule zoomed in, and you can't measure it.
PPC and Variance: Measuring Whether You're Getting Better
Here's the metric that makes the whole thing honest. Percent Plan Complete is dead simple: of the activities you committed to this week, how many finished, completely, as promised? Ten commitments, seven done, that's 70% PPC. It is deliberately harsh — 80% done doesn't count, an activity that slipped a day doesn't count. Binary, on purpose.
New teams routinely land in the 50s and are horrified. Don't be. A low PPC isn't a grade on your crews' effort; it's a measurement of your planning reliability. It tells you your commitments weren't real, which usually means your make-ready wasn't done. Mature Last Planner teams live in the 80s. Chasing 100% is a mistake — if you're always hitting 100, you're sandbagging the plan and leaving productivity on the table.
The number itself is useless without the second half: variance. Every failed commitment gets a reason, and you trend those reasons. This is where the money is. If half your misses trace back to "prior trade not complete," your problem is sequencing and handoff discipline. If they're "materials not on site," your problem is procurement and lead times. If it's "changed priorities," someone above you is thrashing the plan and you now have data to take upstairs. Variance analysis turns anecdotes into a fix. Without it, PPC is just a sad number on a whiteboard.
A practical cadence: capture PPC and reasons at the same Monday meeting where you set the new plan. Review last week, learn, then commit forward. Fifteen minutes on variance every week does more for reliability than any amount of replanning.
Where Software Earns Its Keep
You can run all of this with sticky notes and a spreadsheet, and plenty of great supers do. The reason to reach for a tool is that the manual version buries you in transcription. Every week you're copying the lookahead into a weekly plan, redrawing the board, chasing who committed to what, and re-tallying PPC by hand. That overhead is exactly where the discipline dies — when make-ready and variance tracking become chores, they're the first things to get skipped, and then you're back to a wall chart nobody trusts.
A scheduling tool built for short-interval planning should do a few specific things, and it's worth judging any product against this list:
- Show the work by location, not just by date. Construction is spatial. A plan organized by area and floor matches how trades actually collide, and it makes stacking conflicts visible at a glance.
- Carry trade-flow sequences. When you connect the handoff logic once — this trade follows that one through these areas — the tool can flow the sequence forward instead of you redrawing it every week.
- Roll the lookahead into the weekly plan. The three- and six-week windows should be the same living plan at different zooms, not separate documents you reconcile by hand.
- Share cleanly with subs. The plan is only real if the trades see the same picture you do. A superintendent's board and a crew lead's phone should show one truth.
That last point is where a mobile companion matters. The plan you build in the trailer needs to reach the foreman on the deck. LookAheadWall was built around exactly this shape of work — a visual, location-based weekly plan with trade-flow connections that the office and the field share, plus a companion app so crew leaders carry the current week in their pocket. The point of any such tool isn't the software; it's that the bookkeeping gets cheap enough that you keep doing the practice when the job gets busy. That's the moment software either saves Last Planner on your project or the whole thing quietly reverts to a Gantt chart taped to the wall.
How to Actually Start
Don't roll this out project-wide on a Monday and expect PPC in the 80s by Friday. Pick one phase and one honest pull plan. Run a real six-week lookahead with named constraint owners. Hold one weekly work plan where foremen are allowed to say no. Measure PPC and write down the reasons for a month before you judge anything. The number will be ugly and then it will climb, and the climb is the whole point — it means your planning is getting more reliable and your crews are spending less time waiting on somebody else's unfinished work.
The superintendents who get real value out of Last Planner aren't the ones who memorized the framework. They're the ones who do the boring middle — the make-ready and the variance — week after week, and who use a tool that makes that middle cheap enough to sustain. Do that, and the coordination stops being something you fight for and starts being something the plan just does.