Every superintendent gets handed a burning building eventually. Mine came about eight years into the career: a four-story podium job, six weeks behind by the time anyone said the word "behind" out loud, three subs threatening to demobilize, and a wall of Post-it notes in the trailer that hadn't been touched since the last owner meeting. The GC's answer was to add a Saturday and yell louder on the morning call. That doesn't work. You can't out-hustle a coordination problem, and a project in trouble is almost always a coordination problem wearing a schedule problem's clothes.
The Last Planner System is the most reliable way I know to pull a job like that back from the edge. Not because it's a magic framework, but because it forces a discipline the project has clearly lost: making commitments people can actually keep, and dealing honestly with the reasons they can't. Good look-ahead scheduling software makes that discipline easier to run, but the practice is what does the work. Here's how to actually apply it when the job is already on fire.
First, diagnose what "behind" really means
Before you plan a single recovery week, figure out which kind of trouble you have, because the fix is different for each. In my experience troubled jobs fall into three buckets, and most have some of all three:
- A sequencing problem. The work is being attempted in the wrong order. Drywall is chasing MEP rough-in that isn't inspected, or you're painting rooms the flooring crew still has to tear back into. This is the most common and the most fixable.
- A constraint problem. Crews show up ready and get turned away because a submittal isn't approved, material didn't land, a permit's hung up, or the area isn't released by the trade ahead. The schedule isn't slow; it's blocked.
- A capacity problem. A sub genuinely doesn't have the manpower to hit the durations, and no amount of planning conjures bodies that don't exist.
Spend the first few days walking the job and sitting in on the huddle before you touch the schedule. Count how many planned tasks last week actually finished. If it's below half, you don't have a manpower problem yet, you have a reliability problem, and reliability is what Last Planner fixes.
Stabilize before you accelerate
The instinct on a recovery is to slam a compressed schedule on the wall and demand everyone hit it. Resist that. A compressed schedule full of tasks that will slip is just a slower way to fail, and it burns the crews' trust the first week when nothing lands.
Start narrow. Run one honest weekly work plan. Get the foremen in a room, look out at the next five working days, and make commitments only for work that is genuinely ready to go, meaning the area's released, the material's on site, the crew's assigned, and the trade ahead is clear. That "make-ready" test is the whole game. If a task fails any part of it, it doesn't belong in this week's plan no matter how far behind you are.
Then, and this is the part everyone skips, measure it. At the end of the week, count the tasks you committed to versus the ones that actually completed. That percentage is your Percent Plan Complete, and on a troubled job the first honest number is usually somewhere between 35 and 55 percent. That number stings, and it should. It's the first true thing anyone has said about the project in weeks. Track it every week from there. Watching it climb toward 80 is how you'll know the recovery is real and not just optimism.
A tool like LookAheadWall earns its keep here because it puts the weekly plan, the trade-flow sequence, and the PPC trend in one place the whole team can see. But even a whiteboard works for the first week. The medium matters far less than the honesty.
Pull the constraint log out of everyone's head
On a healthy job, the look-ahead window is where constraints get killed before they ever hit the work plan. On a troubled job, that window has usually collapsed to about three days and everybody's living in the emergency. Rebuild it deliberately.
Reach out to a two-to-three-week look-ahead first, not six. Six weeks sounds thorough, but on a project that can't reliably predict next Tuesday, a six-week plan is fiction, and fiction teaches the crews to ignore the schedule. Take every task in that near window and ask one question: what has to be true for this to start on time? Submittal approved. Long-lead material confirmed with a real ship date, not a hopeful one. Inspection scheduled. Predecessor trade complete and signed off. Each of those becomes a line in a constraint log with a name and a date attached, and somebody owns clearing it.
This is where most recoveries actually turn. The schedule slippage was never the disease; it was the fever. The disease is a stack of unaddressed constraints, and until you make them visible with an owner and a due date, they keep resetting your plan every week. Constraint management software gives you a running list that doesn't live in one person's memory, so when a foreman leaves the job the log doesn't leave with him.
Fix the sequence, don't just resource-load the mess
Once the plan is reliable and the constraints are visible, look hard at the trade flow. Recovery is your chance to re-establish a clean sequence, because sloppy handoffs are usually what got the job here. A few sequencing habits that save recoveries:
- Build a real buffer between rough-in and cover-up. Frame-to-rough-in wants a day or two for cleanup, corrections, and inspection before drywall closes anything. Skip it and you'll be cutting inspection holes in finished board, which is slower than the buffer you tried to save.
- Megger and pressure-test before the wall closes, not after. Every run you close blind is a bet you'll pay off later at ten times the cost.
- Run trades through areas in a consistent train, same crews hitting the same zones in the same order, so a foreman can predict when his area frees up. Random room-hopping is how you get three trades tripping over each other in the same corridor.
- Protect finished work. On a recovery especially, don't send a crew back into an area you already closed out unless you truly have to. Every re-entry is rework you didn't budget.
Location-based, visual planning helps here because you can see the collisions before they happen, two trades stacked in the same zone in the same week, instead of discovering them at 6 a.m. when both crews show up.
Re-engage the subs, because you need them more than they need you
By the time a job needs recovery, the subcontractor relationships are usually strained. Crews have shown up to blocked areas one too many times and stopped trusting the schedule, so they staff you light and save their good people for jobs that run clean. You have to earn that back, and you earn it by being reliable first.
Get the sub foremen into the weekly planning conversation and let them make their own commitments rather than receiving marching orders. People keep promises they made themselves. And hold up your end visibly: when the GC owns a constraint, put it on the same log with the GC's name on it, and clear it. Nothing rebuilds credibility faster than a super who tracks his own obligations as ruthlessly as he tracks the subs'. A companion app that puts the week's plan in a crew leader's pocket helps too, but it's the follow-through that changes behavior, not the notification.
Compress with a scalpel, not a hammer
You will eventually have to make up time, and Last Planner discipline actually makes that safer. The rule is simple: only accelerate work that's ready. Throwing overtime and extra crews at tasks that are still constrained just spends money to produce the same slip, plus fatigue. Find the true critical path, clear its constraints ahead of everything else, and pour your acceleration there. Everywhere else, hold a sustainable pace. Sixty-hour weeks buy you two good weeks and then a month of injuries, callbacks, and burnout, which on a recovery is the last thing the job can absorb.
What success actually looks like
You'll know the recovery is taking hold before the CPM schedule shows it. The leading indicator is PPC climbing week over week, from the low 50s into the 70s and holding. Behind that number you'll feel the job change: the morning huddle gets shorter because there are fewer surprises, problems start surfacing in the planning meeting instead of at the work face, and subs quietly start bringing more people because the areas are finally ready when you say they will be.
The last thing worth saying is the hardest to hear on a job that's already hurting: recovery is far more expensive than prevention. Rebuilding trust, unwinding rework, and compressing what's left all cost more than running the discipline from day one would have. So take the real lesson forward. The weekly work plan, the make-ready look-ahead, the honest PPC number, and a live constraint log aren't recovery tools. They're how you keep the next job off this list entirely.