Every super has watched a beautiful CPM schedule go stale by the second week of a job. The bar chart says drywall starts Monday, but the framer's still chasing a change order, the inspector hasn't signed off on rough-in, and the material's sitting on a truck in a yard two states away. The master schedule tells you where you're supposed to be. It does a lousy job of telling you what to actually do on Tuesday. That gap is exactly what a four-week lookahead is built to close.
A four-week lookahead is a rolling window you pull off the master schedule and turn into work you can actually commit to — activities broken down small enough to hand a foreman, with the constraints in front of each one flagged and being chased down before the crew shows up. Done right, it's the difference between running the job and reacting to it. Here's how to stand one up so it survives contact with a real jobsite.
Why four weeks and not three or six
The window length isn't arbitrary. Look too far out and the plan is fiction — you're guessing at durations and sequences that will change three times before you get there, and crews learn to ignore it. Look too close in and you don't leave enough runway to remove constraints. If a submittal takes three weeks to approve, a two-week window means you're already late the day you spot the problem.
Four weeks is the sweet spot for most commercial and multi-family work because it roughly matches the lead time on the constraints you can still do something about. Long-lead equipment and major submittals need their own tracking further out, but the day-to-day chokepoints — inspections, a missing detail, a sub who needs a week's notice to mobilize a second crew, material that ships in ten days — all live inside a four-week horizon. You can see them coming and still have time to clear them.
Treat the four weeks as a gradient, not four equal blocks. Week one should be nearly locked — the work is planned to the crew and the day, constraints are gone, and if something's still open in week one you have a problem. Weeks two through four get progressively looser. By week four you're mostly confirming that constraints are being worked, not committing to a specific Thursday. Each week the whole window rolls forward one week, week two becomes the new week one, and you tighten it up.
Breaking activities down to the right size
The single most common mistake is dragging master-schedule activities straight into the lookahead. "Level 3 Framing — 15 days" is a planning object, not a work assignment. Nobody can commit to it and nobody can tell you Wednesday afternoon whether it's on track.
Break work down to activities a single crew can finish in one to five days, tied to a physical location. Instead of "Level 3 Framing," you want "Frame demising walls, units 301–308," "Frame corridor east wing," "Set door bucks Level 3." Now a foreman can look at it, tell you whether it's realistic, and tell you at the end of the day whether it got done. This is where location-based, visual planning earns its keep — when activities are pinned to areas rather than buried in a list, the crew can see the flow of trades through a space and you catch the collisions before they happen. It's the core of what LookAheadWall was built to do, and it's also just good practice with a whiteboard if that's what you've got.
A rule of thumb: if you can't name the crew and the location, the activity is too big. Break it down until you can.
Constraint tracking is the whole game
Here's the part that separates a lookahead from a prettier bar chart: every activity that's going to be made ready has constraints in front of it, and your job is to find them and kill them before the crew is standing there. An activity isn't "ready" because it's next in the sequence. It's ready when the answer to every one of these is yes:
- Prerequisite work — is the preceding trade actually complete in that location, not just "mostly"?
- Materials — is it on site, or confirmed to arrive with buffer, not the morning of?
- Information — are the RFIs answered and the details resolved? A wall you can't dimension isn't ready to frame.
- Approvals and inspections — is the prior inspection signed off? Do you need one before you can proceed?
- Labor and equipment — does the sub have the crew size committed, and is the lift or the scaffold available?
- Access and conditions — is the area handed over, safe, and physically reachable?
Run that checklist against every activity in weeks two through four. Anything with an open constraint gets an owner and a date. That's the actual work of managing a lookahead — not drawing the bars, but working the constraint log so that by the time an activity hits week one, it's clean. Crews that consistently hit their weekly commitments aren't luckier than everyone else; they just don't let work into the plan until the constraints are gone. That discipline, making work ready before you commit to it, is the heart of the Last Planner approach and it's what makes a lookahead more than decoration.
The weekly cycle that keeps it alive
A lookahead is not a document, it's a habit. Pick a day — a lot of jobs run it Thursday or Friday so the coming week is locked before the weekend — and hold the meeting every single week without exception. The rhythm matters more than any one meeting.
Keep the foremen in the room. Not their office. This is a planning session run by the people who'll do the work, and it should be blunt and fast — usually 45 minutes to an hour if you're disciplined. A workable structure:
- Score last week first. Of the activities you committed to in week one, how many actually got done? Track that percentage. When you miss, don't move on — ask why, and write the reason down. Patterns show up fast: the same sub over-promising, materials always late from one supplier, inspections you keep forgetting to call for.
- Lock the new week one. Walk the coming week activity by activity. Each foreman confirms — out loud — that the work is ready and they're committing to it. If a constraint's still open, it doesn't go in. Move it out and find something that is ready.
- Work weeks two through four. Look ahead at what's coming and hunt for constraints. Assign each open one an owner and a due date. This is where you buy yourself the runway to not be firefighting next week.
- Roll it forward. Advance the window and set up for next week.
That "why did we miss" number — Percent Plan Complete, if you want the formal name — is the most honest metric on the job. A crew hitting 45% isn't lazy; the planning is broken upstream and the reasons will tell you exactly where. A crew consistently above 80% is a well-run operation. Watch the trend more than any single week.
Who owns it, and the failure modes to watch
The superintendent owns the lookahead. Not the project engineer, not the scheduler in the office. The person accountable for getting work put in place has to own the plan for putting it in place, because the whole point is committing to work you can actually deliver. The PE can help maintain it and chase constraints; the super drives it.
A few ways these die, and how to keep yours breathing:
- It becomes a report instead of a plan. If you're building it Friday afternoon to show the owner where you are, it's dead. It has to drive next week's work or it's theater.
- Foremen aren't in the room. A lookahead built by management and handed down gets no buy-in and no honest push-back. The people committing to the work have to build it.
- Constraints get logged but not chased. A constraint list nobody works is just a nicely organized record of why you're going to be late. Every open item needs an owner and a date, and you follow up.
- You skip a week "because it's slow." The week you skip is the week the plan stops being trusted. Hold it every week, even a short one.
- Activities are too big to commit to. If foremen can't honestly say yes or no to an activity, it's not broken down enough. Fix the granularity.
Whether you run it on a whiteboard, a spreadsheet, or a purpose-built tool matters less than the discipline behind it — though once you're sharing the plan with a dozen subs and rolling it every week, software that keeps everyone on the same live view and flags the constraints saves real headaches. LookAheadWall handles that sharing and the visual trade-flow side, but the practice comes first. Get the four-week habit right, keep the foremen honest and the constraint log worked, and you'll spend a lot less of your day reacting to the job and a lot more of it running it.