Most Last Planner System rollouts don't fail because the methodology is wrong. They fail because somebody bought software, held a kickoff lunch, and then expected the crews to change how they've planned work for thirty years by Friday. Six weeks later the whiteboard is back up, the software login is forgotten, and everyone quietly agrees LPS "doesn't work on our kind of job."
It does work. I've watched a badly behind hospital job claw back a month of float once the trades started actually committing to what they could hold. But getting there is a sequence, and like any sequence on a jobsite, doing the steps out of order or skipping the make-ready gets you a mess. Here's how to stand up the Last Planner System with software backing it, in an order that actually holds.
Get the leadership problem out of the way first
Before you touch a tool, be honest about one thing: LPS asks your project team to make public commitments and then measure how often they keep them. That is threatening to a lot of people, especially superintendents who've survived their whole careers by keeping the schedule in their head and never being pinned down.
So leadership buy-in isn't a slide. It's the PM and the super agreeing, out loud, that a missed commitment is a learning event, not a firing offense. The first time a foreman says "I can't commit to that pour, the rebar inspection isn't signed off" and gets thanked instead of chewed out, you've won more than any training session will buy you. If your culture punishes honest "no," people will just keep saying "yes" and lying to the schedule. The software will faithfully record those lies.
Teach the method before you teach the buttons
The single most common rollout mistake is running a software training and calling it an LPS training. They are not the same thing. Your team needs to understand four ideas before a screen helps them:
- SHOULD, CAN, WILL, DID. The master schedule says what you should do. The look-ahead screens for what you can do once constraints are cleared. The weekly work plan is what the crews will do. And you measure what actually got done. Every screen in the tool maps to one of those.
- Make-ready is the whole point. The look-ahead window exists to find and kill constraints before the work is due — not to be a prettier bar chart.
- Only the last planner commits. The foreman or crew leader who will actually do the work makes the commitment. Not the PM on their behalf.
- PPC measures the plan, not the people. Percent Plan Complete tells you whether your planning is reliable, so you can fix the system.
Spend a couple of hours on this with the whole team, trades included. If they get the "why," the software becomes obvious. If they don't, no amount of clean UI saves you.
Pick the software, and pick it for the make-ready
When you evaluate a tool, the tempting thing is to score it on how pretty the Gantt looks. Ignore that. The Gantt is your master schedule and you probably already have one in P6 or MSP. What you're buying LPS software to do is the stuff those tools do badly: rolling look-aheads, constraint tracking, weekly commitments, and PPC.
Score candidates on questions that actually bite:
- Can a foreman flag a constraint on an activity from their phone, standing in the field, in under fifteen seconds? If it takes a laptop and a login, it won't happen.
- Does the weekly work plan read the way crews think — by location and by crew — or only by CSI activity code?
- Can subs see and update their own commitments without you re-keying their promises?
- Does it calculate PPC for you, or are you going to be doing it in a spreadsheet at 9pm on Thursday?
A visual, location-based look-ahead tool like LookAheadWall earns its keep here precisely because it's built around the weekly work plan and the trade-flow sequence, not around retrofitting a critical-path chart. But whatever you choose, the test is the same: does it make the make-ready process easier, or does it just digitize the whiteboard?
Choose a pilot you can actually win
Don't pilot on your worst job hoping LPS rescues it. A troubled project has too many variables to teach you anything clean. Pick a job with a receptive super, a couple of trades you have a good relationship with, and at least three or four months left to run — long enough to complete several planning cycles and see PPC trend.
Structural or interior-fit-out phases with lots of trade handoffs are ideal, because that's where reliable commitments pay off fastest. One building, one phase, one honest scorecard. Prove it there, then let the results sell the next job for you.
Train your facilitator like it's a trade
The weekly and phase-planning meetings live or die on facilitation. A good facilitator keeps the room on commitments and off war stories, makes sure the actual last planner is the one talking, and refuses to let a "we'll figure it out" pass for a real plan. That's a learned skill. Pick one person — often the super or a project engineer with credibility in the field — and let them run several sessions before you judge the process.
Start with pull planning, working backward from a milestone
Here's where the real work begins. Get the trades in a room (or around the same board) for the phase you're planning, and build the sequence backward from a milestone — "dried-in," "permanent power," "TCO." Each trade puts up what they need to hand off to the next trade and how long their piece takes.
You'll surface conflicts you'd otherwise hit in the field: the drywaller assuming overhead MEP is done, the flooring sub needing the slab cured longer than the schedule allows. Working backward forces the handoffs into the open. Capture that pull plan in the software so it becomes the spine your look-ahead hangs on — this is your trade-flow sequence, and it's the thing most schedules never write down.
Build the make-ready process, because this is the engine
If you take one thing from this whole guide, take this: the make-ready process is the Last Planner System. Everything else is scaffolding around it.
A rolling look-ahead — most jobs run a three-to-six-week window — exists so that for every activity coming due, someone screens it for constraints and clears them before the crew shows up. The usual suspects:
- Prerequisite work not complete (the handoff isn't ready)
- Materials not on site or not released
- Missing information — an RFI still open, a detail not resolved
- Inspection or sign-off not scheduled
- Equipment, crane time, or access not arranged
- Permit or engineering hold
Every constraint gets an owner and a need-by date, and it gets tracked until it's gone. The rule of thumb: an activity doesn't earn a spot on the weekly work plan until it's constraint-free. That single discipline — never committing to work that isn't ready — is what makes PPC climb. Let the field flag new constraints from their phones, because half of them only become visible when someone's standing in the room looking at the actual condition.
Launch weekly work planning as a real commitment ritual
Now the weekly meeting means something. You're only pulling activities that survived make-ready, so the crews are committing to work that's genuinely ready to go. Keep it short and make it about promises: each last planner looks at the coming week and says what their crew will do — not what they hope to, not what the master schedule wishes they would.
Load those commitments into the tool by crew and location so the foreman can pull up exactly their week on a phone the next morning. A weekly plan the crew can actually see is a weekly plan the crew actually runs. Leave a little buffer between trades on the handoffs — a day or two on something like frame-to-rough-in for cleanup, punch, and inspection saves you the cascade when one trade slips.
Track PPC, and treat the misses as gold
At week's end, mark each commitment done or not done. Percent Plan Complete is simply the fraction you kept. Most teams start somewhere in the 50s and, done right, work into the 80s over a few months.
But the number itself isn't the prize — the reasons for the misses are. Every "not done" carries a variance reason: waiting on RFI, materials late, prerequisite slipped, weather, crew pulled to another task. Tally those reasons and the pattern jumps out. If a third of your misses are "prerequisite not complete," your make-ready screening is too shallow. If it's "materials late," your procurement lead times are wrong. PPC without root-cause analysis is just a bad grade; PPC with it is a to-do list for fixing your planning system.
Make the learning loop stick
The first few weeks will feel clumsy and your PPC will be humbling. That's normal — you're measuring reality honestly, probably for the first time. Keep the reason codes visible, keep closing the loop on the top recurring cause, and let the crews see the number climb. Nothing sells LPS to a skeptical foreman like watching his own reliability go up because the work is finally ready when he gets to it.
The software's job through all of this is to carry the load you don't want to do by hand — rolling the look-ahead forward, holding the constraint log, capturing commitments in the field, and doing the PPC math — so your people spend their energy on the conversations that actually move the job. Get the sequence right, protect the honest "no," and the system does the rest.