I've watched three Last Planner rollouts go cold. Same story every time: a kickoff meeting with donuts, a slick piece of scheduling software, everybody nodding along about pull planning and percent plan complete. Six weeks later the sticky notes are gone, the weekly work plan meeting has quietly turned back into a status update, and the software login gets used once a month to print something for the owner. Nobody killed it. It just starved.
The thing that was missing wasn't training and it wasn't the tool. It was a person. Somebody who owned the practice, showed up every week whether they felt like it or not, and dragged the rest of the team along until the habit set. On the jobs where Last Planner stuck, that person was always there. On the jobs where it died, there was a committee and no owner.
That's what this article is about. Not the theory of a "champion" — the actual job of being one, why the software makes it harder and easier at the same time, and how to keep the whole thing from collapsing the day that person gets pulled to another project.
Why the software makes the champion more important, not less
You'd think a good tool would reduce your dependence on any one person. In practice, look-ahead software raises the stakes. A wall covered in Post-its is visibly falling apart when nobody maintains it — the notes curl, they go stale, the superintendent walks by and feels the guilt. Software hides that decay. The schedule still looks clean on the screen even when nobody's touched the commitments in two weeks. The failure is silent, and silent failures are the ones that win.
Last Planner is also a change in behavior before it's a change in tooling. You're asking a plumbing foreman to publicly commit to finishing an area by Thursday, in front of the drywall foreman who's waiting on him, and then to stand there next week and say out loud that he didn't make it. That's uncomfortable. People don't adopt discomfort because a training video told them to. They adopt it because someone they respect keeps the ritual honest, week after week, until the discomfort turns into normal. Tools don't do that. People do.
What a real champion actually does on Tuesday at 2 p.m.
Forget the personality-trait checklists. Here's the work, concretely.
- Runs the weekly work plan meeting on time, every time. Same day, same hour, 45 minutes, hard stop. When the champion lets it slip to "whenever we can grab everyone," it's already dying. Protecting that slot on the calendar is half the job.
- Screens the six-week look-ahead for constraints before the meeting, not during it. Walks the upcoming activities and asks the boring questions: Are the submittals approved? Is the material on site or has a confirmed delivery date? Is the preceding trade actually done, or just "mostly"? Is the inspection scheduled? An activity with an open constraint doesn't belong in next week's plan, and the champion is the one who catches it upstream.
- Makes foremen commit in their own words. Not "we'll try to get to the third floor." A specific, measurable, done-or-not-done commitment: "east wing, gridline 4 to 8, ready for inspection Thursday." Vague commitments are how PPC gets gamed.
- Measures Percent Plan Complete and posts it where people see it. PPC is just: commitments met divided by commitments made. If your team is running above 90% consistently, they're sandbagging — planning only what they already knew they'd finish. Healthy PPC on a real job lives around 65–80% while you're still learning, climbing as the planning gets tighter. The number matters less than the conversation it forces.
- Runs the five-whys on every miss. This is the part almost everyone skips, and it's the whole point. When a commitment isn't met, you capture the reason for variance. Do it long enough and the same three or four root causes show up over and over — usually prerequisite work not finished, a late material, or an over-optimistic duration. Fix those and PPC climbs on its own.
None of that requires software. But once you're running it across multiple areas and trades, a shared look-ahead tool is what keeps everyone looking at the same plan instead of five private versions. This is exactly the spot where a visual, location-based tool like LookAheadWall earns its keep — the constraint log, the trade-flow sequence, and the PPC tracking live in one place your subs can actually see, and the crew leads can pull it up on their phones in the field instead of squinting at a printout from Monday. The tool doesn't run the meeting. It just makes the champion's job less of a paperwork slog.
You don't need one champion. You need three kinds.
The most common mistake is pinning everything on one enthusiastic superintendent and calling it done. Last Planner needs coverage at three altitudes, and they do different jobs.
The executive who protects the time
This person rarely touches the schedule. Their job is political: making sure the weekly planning process is treated as real work and not overhead to be cut the first time a project runs hot. When a PM tries to skip the meeting because "we're too busy this week," the executive champion is the one who says no. Without air cover from above, the field champion eventually loses the argument to schedule pressure — because they always do, on their own.
The field champion who runs the ritual
Usually the superintendent or a strong lead. This is the person doing the Tuesday-at-2 work above. They have to have real credibility with the trades — a champion the foremen don't respect is worse than no champion, because now the process is associated with someone they tune out. Pick the person the subs already listen to on site, then teach them Last Planner. Don't do it the other way around.
The technical champion who keeps the tool honest
Somebody has to own the software itself: set up the projects, keep the trade flows current, fix it when the constraint log gets messy, and train the next foreman who rotates on. On a small job this can be the same person as the field champion. On anything bigger, split it — a great superintendent who hates computers will let the tool rot, and then everyone blames the software.
How to find them (hint: it's not who volunteers)
The loudest volunteer in the kickoff is often the worst pick — enthusiasm burns off fast when the work turns into a weekly grind. Look instead for the foreman who already plans two weeks ahead in his head, who complains about not getting clean handoffs from the preceding trade. That complaint means he already thinks in flow. He's a champion who doesn't know it yet.
The traits that actually predict a good champion: peer credibility (do the other trades listen to them?), follow-through on boring tasks, and enough job security to say an uncomfortable thing in a meeting without worrying about it. Curiosity about the method helps, but I'd take a stubborn, respected foreman who's lukewarm on the theory over an excited newcomer nobody knows.
The failure modes, and how to see them coming
Rollouts don't die suddenly. They erode in patterns you can spot early:
- The meeting drifts back into a status report. Symptom: people talk about what happened last week and never make specific commitments for next week. The forward-looking commitment is the entire engine — lose it and you're just having a slower progress meeting. Fix: the champion cuts off the retrospective and forces the question "what will you finish, specifically, by when?"
- PPC goes suspiciously high. If everyone's hitting 95%+, they've learned to commit only to sure things. You've traded honesty for a pretty number. Fix: push the planning window out and ask for more ambitious commitments; a little healthy failure is the sign it's working.
- Constraints get logged and never cleared. The constraint list becomes a graveyard nobody works. Fix: every open constraint needs a name and a date attached, and the champion chases them between meetings — that chase is the real value of the whole system.
- The champion becomes a single point of failure. Everything runs through one person, and the day they take vacation the meeting doesn't happen. This is the quiet killer. It's also completely preventable, which brings us to the last point.
Build the thing to survive losing your best person
Here's the hard truth after twenty years: the champion who makes it work is also the reason it's fragile. If the practice only lives in one head, it leaves when that head gets promoted or moved to the problem job across town. Plan for that from day one.
Cross-train a second person on the meeting facilitation early — have them run it once a month while the primary champion sits in the room and stays quiet. Document the ritual in one page: when the meeting happens, what gets reviewed, how commitments are worded, how PPC is calculated. Keep the look-ahead, the constraint log, and the variance reasons in a shared tool rather than in the champion's notebook, precisely so the knowledge doesn't walk out the door with them. And when a champion is clearly heading for burnout — you'll see it, the energy goes flat — rotate them into a mentoring role and bring up the next one before they quit the whole thing in frustration.
A short-interval planning practice that depends on a single hero isn't a system. It's a personality. Systems survive people leaving. That's the difference between a rollout that's still running three jobs later and one that's a fond memory from that one project where things clicked.
The bottom line
Software will give you a cleaner look-ahead, a shared view for your subs, and PPC math you don't have to do by hand. What it won't do is show up Tuesday at 2, force a plumbing foreman to make an honest commitment, and chase down the reason last week's drywall slipped. That's a person's job — and until that person is named, backed by someone above them, and paired with a successor who can carry it, you don't have a Last Planner implementation. You have a login.
Pick the right people. Give them the time and the cover to do the work. Then build the practice so it outlives any one of them. Do that, and the software finally does what the brochure promised.