Every superintendent has lived the Monday that fell apart by nine. The framers show up ready to close walls, but the electrician never got his rough-in signed off because the inspector red-tagged a missing nail plate on Friday afternoon and nobody caught it. Now you've got a framing crew standing around, a sub asking where to go, and a week of sequence that just slid a day to the right before the coffee got cold.
The plan wasn't wrong. It just didn't ask the one question that separates a schedule from a wish: what could stop this from happening? A weekly work plan that only lists what you hope to accomplish is a to-do list with better formatting. A weekly work plan that names the ways each task can fail, and has an answer ready, is what actually keeps a job moving.
A plan is a set of promises, and promises have preconditions
Think of every activity on your look-ahead as a promise a crew is making. "We'll hang drywall on the east corridor Tuesday." That promise is only good if a stack of things are already true: the rough-in is inspected, insulation is in and signed, the board is on site, the crew is scheduled, and nobody's using that corridor as a lay-down yard. Miss any one and the promise breaks.
The discipline of anticipatory planning is just making those preconditions explicit before the week starts, instead of discovering them one at a time as they blow up. In Last Planner terms these are constraints, and the job of the weekly plan is to only commit work that's genuinely constraint-free. Everything else stays in the look-ahead as a candidate, not a commitment. If you can't honestly say the activity is ready, don't promise it. Making the plan is the easy part; making the plan honest is where the value is.
Where weeks actually die: the five usual suspects
Over enough jobs you learn that problems aren't infinite. Most of what wrecks a week falls into a handful of buckets, and once you know them you can screen for them fast:
- The predecessor didn't finish. The single most common killer. The trade ahead of you ran long, or finished but didn't pass inspection. Your Tuesday work quietly depends on their Monday work, and Monday's already shaky.
- Material isn't here — or it's the wrong material. A confirmed PO is not a confirmed delivery. "It shipped" is not "it's on site and it's the right spec." Long-lead items and anything with a submittal are your usual heartbreak.
- The crew isn't right. Bodies got pulled to another job, the sub sent three when he promised six, or the guys who showed don't have the certs the task needs.
- Weather. Obvious for exterior work, but people forget it kills interior work too — no crane pick means no material to the deck, a wet slab won't take flooring, and you can't do a blower-door test with the building open.
- Equipment. The lift is double-booked, the pump's down, the man-hoist is being inspected the exact morning you needed to move material up.
None of this is exotic. That's the point. When you review a week, run each committed activity against these five. It takes ten minutes and it catches most of what would otherwise ambush you.
Walk the week backward and ask "what has to be true?"
Here's the practical drill. Before you lock a weekly work plan, take each activity and ask three questions in order:
- What must already be done for this to start? Name the predecessor and its inspection, by trade and location. Not "prior work" — "east wing rough-in electrical, signed off."
- What has to physically be here? Material on site, the right lift staged, the right crew with the right ticket.
- How sure am I about each of those, honestly? This is the one people skip. A constraint you're 95 percent sure about is fine to commit. A constraint you're 60 percent sure about is a risk wearing a commitment's clothing.
Anything that comes back shaky on question three is your work list for the next 48 hours. That's the difference between a superintendent who spends the week reacting and one who spends the tail end of last week making calls so this week doesn't need reacting to. Constraint resolution is boring, unglamorous phone-and-email work, and it's the highest-leverage thing you'll do all week.
Build the buffers where handoffs actually break
Optimists schedule trades bumper to bumper — plumber finishes Friday, tile starts Monday, zero gap. Then the plumber's Friday inspection fails on a re-pipe, and the whole downstream train derails. Realists put buffer where the risk lives, which is almost always at the handoff between trades.
Some rules of thumb worth stealing:
- Frame-to-rough-in usually wants a one- to two-day buffer for cleanup, punch, and inspection scheduling. Inspectors don't come the instant you call.
- Never schedule an activity to start the same day its predecessor's inspection is due. Inspections fail, get rescheduled, or the inspector shows up at 4 p.m. Give it a morning of margin.
- Leave a slack activity in your back pocket. Location-based planning helps here: if the east corridor gets blocked, you want a ready-to-go task in the west stair or a different floor so the crew moves instead of stands. Idle crews are the most expensive thing on a jobsite.
The mistake is spreading a thin buffer evenly across everything, which just makes the whole schedule mushy. Concentrate the margin at the joints, where the actual failures happen, and run tight where the work is genuinely independent.
Don't schedule to a hundred percent
A crew loaded to full theoretical capacity has no room to absorb a problem. The day someone calls in, a delivery's late, or a task takes longer than the estimate, there's no give and the plan tips over. Planning to roughly 80–85 percent of capacity isn't slacking — it's building in the reserve you'll need when reality shows up, because it always does. The same logic applies to material: order for waste and error, because the run you cut wrong at 3 p.m. shouldn't stop you until tomorrow's delivery.
Your history is a forecast — use it
The best predictor of next month's problems is last month's. If your plan-percent-complete keeps tanking on the same trade, or the same kind of task keeps slipping, that's not bad luck — that's a pattern telling you where to put your attention. Track why committed work didn't get done, not just whether it did. A few weeks of honest variance reasons ("waiting on inspection," "material short," "crew pulled") and the recurring failure modes on your specific job jump right out.
This is where a real scheduling tool earns its keep. When your weekly work plans and their outcomes live in one place — as they do in a look-ahead app like LookAheadWall rather than scattered across whiteboards and spreadsheets — you can actually see the variance history instead of relying on memory. Flag the high-risk activities visually so the whole team reads the week the same way, and let the trade-flow connections make a broken predecessor obvious the second it slips. The value isn't the software; it's that the pattern stops hiding.
Your foremen already know what's going to go wrong
The people closest to the work see the problems first, and they're usually happy to tell you — if you ask before the plan is locked instead of after it fails. Bring the concern out in the planning conversation: walk the upcoming week with your foremen and subs and ask, plainly, "what worries you about this?" You'll hear about the tight access nobody drew, the material that's always short, the inspector who's picky about one specific detail. That's free anticipation, and it's the anticipation that matters most because it's grounded in the actual conditions on your job, not a template.
Do the same with subs on their upcoming scope. A sub who tells you Thursday that he's a man short next week just handed you three days to fix it. A sub who tells you Monday morning handed you a problem. The relationship that gets you the early heads-up is worth more than any contingency on paper.
Watch it develop, and escalate before it's a crisis
Anticipation doesn't stop when the week starts. The morning huddle is your early-warning system — a two-minute check on whether today's preconditions actually held overnight. Deliveries that were supposed to land, inspections that were supposed to pass, crews that were supposed to show. Catch the gap at 6:45 a.m. and you can re-sequence before the crew's even geared up. Catch it at 10 and you've already burned the morning.
And when an anticipated problem becomes a real one, escalate fast and escalate with information. "The main switchgear slipped two weeks, here's what it hits and here are the two options" gets a decision. "Something's wrong with the schedule" gets a meeting. Because you thought the failure through in advance, you already have the contingency ready — the backup task, the alternate crew assignment, the second material source. That's the whole payoff: the problem shows up, and instead of a surprise, it's a scenario you already have an answer for.
The habit, not the paperwork
None of this requires a thick binder or a certification. It's a habit of mind you apply every time you build a week: for each thing you're about to promise, name what has to be true, be honest about how sure you are, buffer the handoffs, keep a little reserve, and ask the people doing the work what scares them. Do that consistently and your plans stop being a list of hopes and start being something the crew can actually build against.
The reactive superintendent spends every week putting out fires. The one who plans for problems spends most weeks quietly handling scenarios he already saw coming. Same job, same chaos out there — completely different week. The difference is asking "what could go wrong?" while you still have time to do something about the answer.