The Last Planner System is one of those ideas that sounds almost too simple when you hear it: the people who do the work should plan the work, and you only commit to things that are actually ready to go. Run it on a 12,000 square foot dental office fit-out and it can be a fifteen-minute conversation at the tailgate. Run it on a 60-story tower or a $900M hospital campus and it becomes a machine with dedicated staff, war rooms, and a metrics dashboard. Same DNA, wildly different body.
Most of the pain I've watched crews go through with Last Planner comes from getting the scale wrong. A super drowns a bathroom remodel in whiteboards and PPC charts, everyone rolls their eyes, and the whole thing dies. Or a project team tries to run a $200M job off the same casual "we'll sort it out Monday" habits that worked on their strip-mall jobs, and by month four the trades are stacked on top of each other and nobody can tell you what's actually ready. This article is about matching the weight of the process to the weight of the job.
What stays the same at every size
Before we talk about scaling anything, it's worth being clear about what doesn't scale — the parts you keep no matter how big or small the job is:
- The people who do the work plan the work. Not the office. Not a scheduler who's never walked the deck. The foreman who's going to hang the drywall tells you when it can go up and what he needs first.
- You don't put anything on the weekly plan that isn't ready. No missing material, no unanswered RFI, no incomplete predecessor. If it's not ready, it stays on the lookahead as a constraint to clear — it does not become a commitment.
- A commitment is a promise, not a wish. When a trade says "done Thursday," that's a reliable promise you can build the next guy's plan on top of.
- You measure whether the promises came true and you learn from the misses. That's Percent Plan Complete (PPC) and the variance reasons behind it. Track why things slip, not just that they slipped.
Everything below — the meetings, the horizon, the software, the paperwork — is just packaging around those four ideas. Scale the packaging. Never dilute the core.
Small projects (under ~$1M): keep it in your head and on one board
On a tenant improvement, a single-family custom home, or a small renovation, the whole trade stack fits in one room, and half of them are guys you've worked with for years. Formality here is the enemy. If you schedule three separate planning ceremonies for a six-week TI, your subs will quietly start skipping them.
What works: one weekly conversation, often standing up, that does double duty as your lookahead review and your weekly work plan. Walk the next two to three weeks, name the constraints out loud — "Is the electrical gear here yet? Did the mechanical drawings get answered?" — and lock the coming week. A three-week lookahead is usually plenty; the job just isn't long enough to see further with any confidence.
You can absolutely run this on a whiteboard or a shared spreadsheet. The value of a proper scheduling tool at this size is mostly that the plan survives past the meeting and the subs can see it on their phones instead of relying on their memory of what you said. Don't over-engineer it. If a fifteen-minute chat and a photo of the board texted to three foremen keeps everyone aligned, you're done.
Medium projects (~$1M–$50M): the textbook version
This is where Last Planner earns its keep and where the "textbook" version lives. A mid-rise apartment building, a school, a decent-sized commercial job — enough trades and enough duration that you can't hold the whole thing in your head anymore, but still one coherent site you can walk in a morning.
Run the full cadence:
- Phase pull planning at the start of each major phase (structure, envelope, rough-in, finishes). Get the trades in a room with sticky notes and pull the sequence backward from a milestone. This is where you catch the sequencing fights before they cost you money.
- A rolling lookahead — typically four to six weeks out — that you review weekly and use to hunt constraints. Every activity that's coming should have someone assigned to make it ready.
- A weekly work plan meeting where trades commit to the coming week's work, and you close out last week's PPC. Protect this time. When the super lets the weekly meeting get bumped for a walkthrough, the discipline erodes fast.
A real rule of thumb at this scale: give yourself a buffer between trades that fight over the same space or hand off dirty work. Frame-to-rough-in usually wants a day or two of slack for cleanup, inspection, and the inevitable "we found something in the wall." Don't butt-joint every activity end-to-end on the plan — a schedule with zero float between trades looks efficient and behaves like a car pileup.
Large projects (~$50M–$500M): plan by area, not by whole
Once a job is big enough that you can't walk it and hold it in one plan, you stop planning the project and start planning areas. A large hospital, a full city block, a data center — you break it into zones, floors, or phases, and each one effectively runs its own Last Planner cycle.
The shift is organizational more than conceptual. You'll have area-based weekly meetings instead of one all-hands, often a dedicated planner or field engineer whose actual job is chasing constraints, and PPC tracked both per-area and rolled up so the project team can see which zone is dragging. Constraint removal becomes a real process with owners and due dates, not something the super does in his head between fire drills.
The classic failure here is under-structuring — running a $150M job on medium-project habits. The symptom is always the same: trades stacking in the same area, the lookahead full of activities nobody actually made ready, and a weekly plan that's really just a wish list. If your PPC is sitting in the 50s and drifting, that's usually not a people problem. It's a make-ready problem — you're committing to work that was never truly ready, and this is exactly the scale at which that catches up with you.
Mega-projects (over ~$500M): a hierarchy of plans
At the very top end — big infrastructure, campus builds, anything with thousands of activities and multiple prime contractors — Last Planner becomes hierarchical. Each zone or work package runs its own full cycle with near-independence, and above them sits an integration layer whose whole job is coordinating the seams between zones and managing shared resources: the two tower cranes everyone's fighting over, the single hoist, the concrete pump schedule.
Here you'll have a dedicated planning organization, cascading meetings (zone-level plans feeding an integration-level plan), and metrics that roll up through several tiers. The lookahead horizon often stretches to six weeks or beyond, not because you can predict the field that far out, but because long-lead procurement and interface coordination between packages demand it. Nobody's promising exact daily work six weeks out — but you absolutely need to see the constraints coming from that far.
Scaling the lookahead horizon
People treat "three-week vs. six-week lookahead" like a doctrine question. It isn't. The horizon is a dial, and you set it based on two things:
- Lead times. Your lookahead has to be at least as long as your longest constraint takes to clear. If a critical switchgear order is eight weeks out, a three-week window is useless for catching it — you needed to be tracking that as a make-ready item long before it hit the near-term plan.
- Uncertainty. The more volatile the work — heavy weather exposure, unresolved design, unpredictable existing conditions in a renovation — the shorter and more frequently updated the near-term plan should be. You commit less far out because you can promise less far out.
In practice most jobs run a longer view (six-plus weeks) for procurement and constraints and a tighter view (one to three weeks) for actual commitments. That's not two schedules — it's one rolling lookahead read at two zoom levels.
Where software actually helps as you scale
Small jobs, honestly, don't need much. The return on a dedicated tool climbs steeply as the project grows, because the things that break at scale — visibility across areas, keeping fifteen subcontractors looking at the same current plan, rolling PPC up from zones to the whole job, seeing a trade-flow sequence stall three activities before it becomes a crisis — are exactly the things a whiteboard can't do.
A good look-ahead scheduling tool like LookAheadWall earns its place on medium and larger jobs specifically because it makes the plan visual and location-based: you can see which zone a crew is in, watch the trade-flow hand-offs connect, and share the current week's plan straight to the foremen's phones so the guy in the field is never working off a printout from last Tuesday. That last part matters more than any feature list — the fastest way to kill a weekly work plan is to let the field version and the office version drift apart. The tool's job is to keep them the same. On a small TI you can do that with a group text. On a 40-story tower you cannot, and that's the whole point.
The scaling mistakes that actually hurt
Four mistakes account for most of the failures I've seen:
- Over-engineering the small job. Full ceremonies and metrics on a six-week TI. The trades disengage and you've taught them Last Planner is bureaucratic theater.
- Under-structuring the big job. Running a $200M project on casual habits. It works right up until the trade stacking starts, and then it's already a schedule recovery problem.
- One-size-fits-all across a portfolio. Copying the mega-project playbook onto every job in the company because that's "the standard." Standards should cover the four principles and the metrics you compare on — not force a renovation to run like a hospital.
- Scaling the process but not the culture. This is the quiet killer. You can install every meeting and every dashboard, but if trades still feel safe over-promising and nobody's honest about why work slipped, your PPC is fiction. Make-ready and reliable promising are cultural before they're procedural. Scale the trust, or the paperwork is just decoration.
Match the weight to the job
The core of Last Planner — those who do the work plan the work, only commit to what's ready, keep your promises, learn from the misses — never changes. What changes is how much structure you wrap around it. A small job wants a light touch and a single conversation. A mega-project wants a hierarchy of plans and a team to run them. Read your job honestly, set the horizon to your longest lead time and your level of uncertainty, and let your tools do the work a whiteboard can't. Get the scale right and Last Planner feels like common sense. Get it wrong in either direction and it feels like a burden — which is usually the first sign it's about to fall apart.