A rolling lookahead is not a document you build once and admire. It's a habit. The whole point is that the schedule moves with the job: every week you drop off the week that just finished, roll a fresh week onto the far end, and fix the middle so it matches what actually happened on the ground. Done right, it becomes the heartbeat of the jobsite. Done wrong, it turns into a stale wall poster that everybody walks past and nobody believes.
This is the quick start I'd hand a foreman or an assistant super who's about to run their first rolling window. No theory dump. Just how to stand one up, how to keep it alive week to week, and where these things usually die.
What "rolling" actually means
Your master schedule ("CPM") is the whole movie — bid to closeout, thousands of activities, mostly wrong the moment it's printed. A lookahead is a short, high-resolution slice of that movie: the next three to six weeks, broken down to the level of detail a crew can actually execute. Rolling just means the slice slides forward on a fixed cadence so the near term is always in focus.
Pick your horizon and hold it. Three weeks is tight and honest — good for fast interiors and finishes where things change daily. Six weeks buys you lead time to chase long-lead material and permits, but the back end gets fuzzy and people stop trusting it. Four weeks is the sweet spot on most commercial and multi-family work: near enough to be real, far enough to see the constraint coming. Whatever you pick, the window length never changes. You're not extending it when you fall behind. You roll it.
Build your first window
Don't try to plan the whole job on day one. Pull the next four weeks off the master, and expand each activity into the tasks a crew would actually recognize. "Level 3 rough-in" on the master becomes, in the lookahead, top-out plumbing east wing, then in-wall electrical east wing, then low-voltage, then framing inspection, then insulation, then hang board. That level of granularity is the difference between a schedule and a wish.
As you lay out those tasks, do three things for each one:
- Sequence it against the trade in front of it and behind it. This is where trade-flow thinking earns its keep — plumbing has to clear the wall before the electrician frames it out, and both have to finish before the framing inspector will sign off and let you insulate. Get the handoffs in the right order now and half your coordination fights disappear.
- Tie it to a location. "Paint" means nothing. "Paint, Level 2, units 201–210" means something a crew can walk to. Location-based planning is what keeps two trades from showing up in the same room on the same morning.
- List its constraints. Every task that can't start yet, and exactly why: material not on site, RFI open, prior inspection not passed, area not released, no manpower committed. Write the reason down. A constraint you can't name is a constraint you won't clear.
A visual, location-based tool like LookAheadWall is built for exactly this — you're placing tasks by area and week and drawing the trade-flow links between them — but the discipline matters more than the tool. Sticky notes on butcher paper have run plenty of good jobs. What can't be skipped is the granularity and the constraints.
The weekly cadence that keeps it alive
A rolling lookahead lives or dies on rhythm. Same day, same time, every week. Most supers I trust run it Thursday or Friday so the plan is locked before the weekend and crews walk in Monday knowing the play. Here's the loop:
- Update actuals first. Walk the job or pull it from your foremen and mark what actually got done — not what was supposed to get done. Green it if it's complete, flag it if it slipped, and be honest about percent complete. This is the step everybody wants to fudge and the one that makes the whole system worthless if you do.
- Roll the window. Drop the completed week, add a new week on the far end pulled from the master. Now you're looking at four fresh weeks again.
- Re-sequence the middle. Whatever slipped pushes everything downstream of it. Fix the handoffs. If drywall slid two days, so did paint, so did flooring — move them, don't pretend.
- Work the constraints. Go task by task through the near weeks and clear or escalate every constraint. This is the real work; more on it below.
- Make commitments. The nearest week — your weekly work plan — is where foremen commit to specific, constraint-free tasks. Only work that is genuinely ready to go should be in there. If it still has an open constraint, it does not belong in the commitment week.
The trap is treating the weekly update as data entry. It's not. It's a planning session. The value is in the conversation about what's blocking week two, not in the tidy grid you print afterward.
Make-ready: the part that separates good supers from clipboard-holders
The reason a lookahead reaches out four weeks and not one is to buy time to clear constraints before a task is supposed to start. This is "make-ready planning," and it's the single highest-value thing the process does.
Every week, for every task in the window, ask the make-ready questions: Is the material on site or confirmed with a delivery date that clears the start? Is the prior work complete and inspected? Is the area released? Is the design question answered — RFI back, submittal approved? Is manpower actually committed, not just assumed?
Rule of thumb: a task should be fully constraint-free by the time it enters your commitment week. If you're clearing constraints the morning work is supposed to start, your lookahead is too short or you're not working it. And watch your buffers between trades — frame-to-rough-in usually wants a day or two for cleanup and inspection before the next trade walks in, and inspection turnaround is a constraint of its own. Assuming the inspector shows up same-day is how a Friday pour becomes a Monday pour.
Integrate progress honestly, or don't bother
The rolling schedule is only as good as the actuals you feed it. If a foreman marks tasks complete that are 80% there, the whole downstream sequence is built on a lie and you'll find out the hard way when the flooring crew shows up to a floor that isn't ready.
Make percent-complete a walking-the-job number, not a conference-room guess. Better yet, close the loop with the field directly — the reason there's a mobile companion app for crew leaders is so the guy who actually did the work is the one reporting it, from the floor, instead of it getting laundered through three people by Friday. However you collect it, the standard is the same: the schedule reflects the building, not the story about the building.
Track variance so you actually get better
Here's the part most crews skip, and it's the part that turns a scheduling chore into a improvement engine. Each week, look at what you committed to in the prior week's plan and count how much of it you actually completed. That's your Percent Plan Complete — commitments met divided by commitments made. Sixty-five percent is a job that's guessing. Eighty-plus is a crew that plans.
The number matters less than the reasons behind the misses. For every task that didn't get done, write down why in a word or two: prerequisite work late, material short, weather, inspection failed, crew pulled to another area, bad estimate. After a month you'll see the same two or three reasons over and over — and now you know exactly what to fix. Maybe your submittals are chronically late. Maybe one sub keeps overcommitting. Variance tracking is how a lookahead stops being paperwork and starts making you money.
Where these things die — and how to keep yours alive
- Too much detail, too far out. Don't plan week four to the hour. It'll all change. Keep the far weeks coarse and sharpen them as they roll forward.
- Skipping a week. Miss one cadence and the schedule drifts from reality; miss two and nobody trusts it again. Protect the meeting like a pour.
- Optimistic actuals. Covered above. The fastest way to kill the whole system is a rosy status report.
- No constraint owner. "Someone should call the supplier" means nobody will. Every open constraint gets a name and a need-by date.
- The super plans alone. If foremen and key subs aren't in the room making their own commitments, it's your schedule, not theirs, and they'll treat it that way. Pull-planning with the people doing the work is the whole game.
Start Monday
You don't need a rollout plan or a month of training to begin. Pull the next four weeks off your master this week. Break them into tasks a crew would recognize, tie each to a location, and write down every constraint you can see. Meet Friday, update actuals, roll the window, clear what you can, and commit the near week. Do that four weeks running and it stops feeling like an extra task and starts feeling like the way you run the job.
The superintendents who swear by rolling lookaheads didn't get there by reading about them. They got there by running the loop until it became muscle memory. Start rough, stay honest, and let the weekly rhythm do the work.