Almost every delay I have ever chased down started weeks before it showed up on the schedule. The drywall crew stood around on Monday because the framing inspection hadn't been signed off, because the framer left a section open waiting on a plumber, because the plumber was waiting on a fixture rough-in dimension that lived in an RFI nobody had answered. By the time it hit the field it looked like a drywall problem. It was really a three-week-old paperwork problem wearing a drywall costume.
That is the whole case for a rolling look-ahead. Not that it magically makes trades faster, but that it drags tomorrow's problems into today's meeting where you still have room to fix them. Below is how that actually works on a live job, and where the usual attempts fall apart.
Delays are almost never a surprise. They're a missed warning.
If you sit with a genuinely delayed activity and walk backward, you land on the same handful of root causes over and over. It's worth naming them plainly, because a look-ahead is only useful if it's hunting for these specific things.
- Something wasn't ready and nobody checked. The task was on the bar chart, so a crew showed up, and the predecessor work, the material, or the deck wasn't actually there. The schedule said "should," and everyone treated it as "can."
- An answer got asked for too late. The RFI or the submittal or the owner decision that gates the work went in the week the work was supposed to start. RFI turnaround on most commercial jobs runs a week or two, and long-lead submittals far longer. Ask on Friday for Monday and you've already lost.
- Two trades wanted the same room. Nobody sequenced the space, so the tile setter and the electrician both planned for the same bathrooms on the same day. One of them loses, and usually both do.
- Material timing drifted off the real sequence. Long-lead items get ordered against the master schedule, then the job moves at a different pace, and the steel or the switchgear or the storefront lands three weeks early (now it's in your way and getting damaged) or a week late (now everything downstream stops).
Every one of these is visible ahead of time if someone is deliberately looking. That someone, and that looking, is the look-ahead.
Why "rolling" is the word that matters
A master schedule is a promise you made at the start of the job. It's necessary, but it doesn't scan for trouble — it just sits there showing dates. A rolling look-ahead is different in one specific way: it moves forward every week and gets re-examined every week, so the plan is always built on what the job actually looks like right now, not what you hoped it would look like at contract signing.
The practical version is a three-to-six-week window that you rebuild each week. Most supers I trust run four weeks. The near weeks carry real detail; the far weeks carry constraints and owners. It's less about the exact number of weeks than about the discipline of dragging the window forward on a fixed cadence and re-asking the same questions.
The tiering usually looks like this:
- Week 1 is a commitment. Named crews, confirmed material on site or staged, specific quantities, specific locations. If it's in week one, someone has personally promised it can happen.
- Week 2 is a plan. Crews penciled, material ordered and tracked, predecessors clearly on track. A couple of loose ends, but a real shape.
- Weeks 3 and 4 are constraints with names on them. You are not scheduling daily work yet. You are listing what has to be true before that work can be committed, and assigning a person to each of those things.
The point of the tiers is that nothing lands in the commitment week cold. By the time an activity is a week out, it has already been looked at three or four times, and every reason it might fail has already had someone chasing it.
Make-ready: the part most teams skip
The engine underneath a good look-ahead is make-ready planning — the plain work of clearing constraints off an activity before it's allowed into the commitment window. This is the step that separates a look-ahead from a prettier Gantt chart. If you only pull dates forward without clearing constraints, you've just repackaged the same optimism.
For every activity heading into the window, run it against a short constraint checklist. Six categories cover almost everything:
- Predecessor work — is the work in front of this actually complete, and complete to the condition this trade needs? "Framed" isn't the same as "framed, backing in, and inspected."
- Material — ordered, confirmed ship date, and a staging spot picked. An unconfirmed delivery date three weeks out is a red flag, not a green light.
- Labor — the right crew, the right size, the right skills, and actually committed by the sub — not assumed because they're "usually around."
- Equipment — lift, hoist, pump, whatever, scheduled with an owner, not a hope.
- Information — every RFI, approval, or shop drawing this task depends on, submitted with a realistic answer date, counted backward from when you need it.
- Space and access — who else is in that area, and how the handoff works. This is where trade stacking gets caught.
A rule of thumb worth living by: an activity does not earn a crew commitment until every one of those boxes is clear. The whole discipline of the Last Planner approach comes down to that one sentence. You commit to what's ready, not to what the master schedule wishes were ready. It feels slow the first month and then it quietly stops the fire drills.
The weekly cycle, in the order it actually runs
Delay prevention is a habit, not a document. The document is just where the habit leaves its marks. A working weekly rhythm looks like this:
- Look back first. What did we commit to last week, and what actually got done? The single most useful number here is Percent Plan Complete — of the tasks you promised, what fraction finished. Not percent of the job. Percent of your promises. A crew running 50% PPC isn't lazy; it's being handed work that wasn't ready.
- Ask why the misses missed. For every incomplete commitment, name the reason — and use the same six constraint categories. When you tally those reasons over a month, the job tells you exactly where it's bleeding. If half your misses are "waiting on information," your RFI process is the problem, not your crews.
- Roll the window forward. New activities enter the far week and get their first constraint pass. Everything else advances a tier.
- Walk the constraints. Every open constraint gets a status and a name. Anything at risk of not clearing in time gets escalated now, while there's still runway.
- Commit next week out loud. The foremen in the room commit to what's ready. Nothing goes into week one that still has an open constraint. If it's not ready, it stays in week two and you find work that is ready to fill the gap.
Miss this meeting once under schedule pressure and it's remarkable how fast the habit rots. The week you're too busy to plan is precisely the week you most needed to.
Where trade coordination quietly gets won
Most stacking and handoff fights die in the look-ahead meeting instead of on the deck — but only if the trades are in the room looking at the same window. When your framer, plumber, and electrician are all staring at the same three weeks, the conflict over who owns the third-floor corridor on Tuesday surfaces as a two-minute conversation instead of a Tuesday-morning standoff with three foremen and idle crews.
Two things make this real. First, handoffs have to be explicit: this trade finishes to this condition, and then that trade starts. "Rough-in is done" is a fight waiting to happen; "rough-in complete, tested, and inspection scheduled" is a handoff. Second, everyone should be able to see everyone else's planned crew loading in a given area, because that's how you catch two subs planning for the same space before they both mobilize.
Buffers, and where to actually put them
A schedule with zero slack isn't aggressive, it's fragile. The trick is putting buffer where the risk actually lives instead of smearing a little padding on every task. A few that earn their keep:
- Leave a day or two between framing complete and rough-in start for cleanup, backing, and the inspection to clear. Trying to inspect and start covering the same wall the same day is how you end up tearing it back open.
- Don't schedule finishes into a space the day the MEP inspection is due. Inspections slip; give the trade behind it a cushion or you'll have painters standing in a room they can't close.
- Protect weather-sensitive work with real windows, and keep a ready indoor backlog so a rained-out pour doesn't idle the whole crew — it just shifts them to work you already made ready.
The couple of numbers to actually track
You can't improve what you don't measure, but you also don't need a dashboard with forty gauges. Two numbers carry most of the weight. Percent Plan Complete tells you whether your commitments are honest — a PPC that lives in the 80s means your make-ready is working; a PPC in the 50s means you're committing to work that isn't ready. And your tally of constraint reasons tells you what to fix, because it points at the one process (information, material, predecessor) that's costing you the most. Everything else is nice to have.
Where the software earns its keep
None of this requires a computer. Plenty of great supers ran it for years with sticky notes on a wall and it worked. What software buys you is that the wall doesn't get lost, the constraints don't fall off it, and the subs who couldn't make the meeting can still see the plan.
A tool built for look-ahead scheduling — LookAheadWall is built around exactly this location-based, trade-flow way of planning — helps most in a few specific spots: it keeps the rolling window and the weekly work plan in one place instead of a stack of stale PDFs, it makes the trade sequences visible to every sub so stacking shows up before mobilization, and it lets a crew leader pull the current plan up on a phone in the field instead of squinting at a printout from Monday. The value isn't the app; it's that the plan stays live and everyone's looking at the same one. Any tool that does that, or a whiteboard you actually maintain, beats a beautiful schedule nobody updates.
How to start without boiling the ocean
If you're standing this up on a job for the first time, keep it embarrassingly simple and let it earn trust:
- Start with a three-week window and only the major activities. You can add resolution once the team has the rhythm.
- Put the foremen and key subs in the room. Constraint identification is impossible from an office — the people doing the work know what's really in the way.
- Hold the meeting on the same day, every week, no exceptions. The cadence is the whole thing.
- Start counting PPC and constraint reasons from week one, even roughly. Within a month the pattern will tell you where your job is actually losing time.
- Keep the tone about the process, not the person. The moment a missed commitment turns into a blame session, people start hiding constraints instead of raising them, and you've lost the early warning you built the whole thing for.
Delays aren't fate. Nearly every one is a warning that showed up weeks early and got walked past. A rolling look-ahead is just the discipline of not walking past it — of asking, every single week, not "what are we supposed to do next" but "what would stop us from doing it, and who's fixing that today." Do that consistently and the fire drills get rare. That's not a software feature. That's a habit, and it's the most reliable schedule-recovery tool you'll ever own.