Walk onto most jobs and you'll find two documents that pretend not to know each other. There's the look-ahead — the three, four, or six-week horizon that lives in a binder or on a screen — and there's the weekly work plan, the list of what crews actually committed to on Monday. On a badly run job, the look-ahead gets built once in the trailer, printed, and slowly goes stale while the real work drifts somewhere else. Nobody reconciles them. By week six the look-ahead is fiction and everybody knows it, so everybody ignores it.
That disconnect is the single biggest reason short-interval scheduling fails to stick. The look-ahead and the weekly work plan aren't two documents. They're two ends of one loop. What you learn executing this week's plan is exactly the fuel that makes next week's look-ahead worth trusting. Miss that feedback and you're not planning — you're just writing wish lists.
How the Two Actually Connect
Think of it as a pipeline with a return line. The look-ahead is your make-ready runway: activities enter the far end (say, week six) rough and full of unknowns, and as they roll toward the near end you scrub the constraints off them — materials ordered, submittals approved, predecessor work complete, area handed over, permit in hand. By the time an activity reaches the front of the look-ahead, it should be clean: no open constraints, ready to hand to a crew.
The weekly work plan is where you pull from that clean pool. You don't put anything into a weekly commitment that still has an open constraint — that's the whole discipline. If it's not ready, it doesn't go on the plan. Then the crews execute, and at the end of the week you measure what actually happened. That measurement is the return line. It flows back and reshapes the look-ahead for the next cycle. Break any segment of that loop and the whole thing quietly dies.
The Number That Runs the Loop: PPC
Percent Plan Complete is the pulse. It's brutally simple: of the activities you committed to this week, how many did you finish completely? Eight of ten commitments done is 80 percent. And "completely" is the operative word — it's all-or-nothing per task, no partial credit. An activity that's "almost done" counts as not done. That strictness is the point — it forces honesty out of a process that's otherwise easy to fool yourself about.
Most crews new to this run 50 to 60 percent PPC and are shocked, because they were sure they were hitting 90. A mature team running disciplined make-ready lands around 80 to 85 percent. You almost never want 100 — if you're hitting 100 every week, you're sandbagging your commitments, planning less than you can actually do. The useful signal isn't the number itself. It's the reasons behind the misses.
Capture the Reason, Not Just the Miss
The mistake I see constantly: a super marks an activity incomplete, rolls it to next week, and moves on. He captured the what and threw away the why. The why is the only thing that improves your planning. Every incomplete commitment needs a variance reason attached, and you want a short, consistent set of buckets so you can count them later:
- Prerequisite work not complete — the trade ahead of you didn't finish, so you couldn't start.
- Materials or equipment — didn't show, showed short, or showed wrong.
- Manpower — crew got pulled, called out, or you underestimated the headcount.
- Information / RFI — waiting on an answer, a detail, or a design decision.
- Access / area not ready — the space wasn't handed over, or another trade was still in it.
- Weather / conditions — the honest ones, not the excuse ones.
- Rework / quality — you had to redo something.
- Bad plan — the commitment was never realistic; own it.
Six weeks of tagged variances turns into a Pareto chart, and that chart tells you the truth about your job. When you see that 40 percent of your misses are "prerequisite work not complete," you don't have a crew problem — you have a handoff problem, and the fix lives upstream in your trade-flow sequencing, not in yelling at Monday's crew.
Updating the Look-Ahead: The Friday-to-Monday Move
Here's the mechanical reconciliation that most people skip. After the weekly review — end of week, before the next planning session — you walk the look-ahead and update it against reality. Four things happen to every near-term activity:
- Completed activities get marked done and cleared. Simple.
- Incomplete activities carry forward — but not blindly. If the reason it missed is still live (still no materials, area still occupied), it stays constrained. Rolling an incomplete task into next week's ready pool without resolving why it stalled is how you guarantee it misses again. That's the number-one self-inflicted wound in this whole system.
- Durations get corrected. If drywall hang took the crew five days and you'd planned three, that's not a one-time miss — that's your estimate being wrong. Fix the duration everywhere that activity type repeats down the schedule, not just this instance.
- New activities enter the far end of the horizon as the window slides forward a week, still rough, ready to start their own make-ready journey.
Do this every week and the look-ahead stays alive. Skip it for even two weeks and it decays into that trailer-wall fiction nobody trusts.
The Make-Ready Reality Check
Your weekly results are also a report card on your make-ready process, and this is where the loop gets genuinely powerful. If an activity was in your "ready" pool — you swore its constraints were clear — and it still couldn't start, then your make-ready missed something. That's a more expensive miss than a task you knew wasn't ready.
So separate your variances by that line. A miss on an activity you knew was risky is a planning honesty problem. A miss on an activity you certified as ready is a constraint-identification problem, and it means a whole category of constraint is slipping past your screening. Maybe you never check whether the inspection got scheduled. Maybe "area handed over" keeps meaning "handed over except the part with the other trade's material stored in it." Find the pattern and push that constraint check further up the look-ahead so you catch it two weeks out instead of Monday morning.
Where This All Happens: The Weekly Planning Meeting
The loop isn't abstract — it lives in one recurring meeting, and the agenda is always the same three moves in this order:
- Look back. Score last week. Read the PPC, and go through every miss with its variance reason. Fifteen minutes, no blame, just facts. The foremen who own the misses out loud are the ones who improve; the ones who make excuses are telling you where next week's misses will be.
- Update the middle. Reconcile the look-ahead against current status — the four moves above. Resolve constraints on activities heading into the next couple weeks. This is the make-ready work, and it's the part everybody wants to shortcut. Don't.
- Look forward. Pull clean activities from the ready pool into next week's commitments — and get the trade foremen to actually commit, out loud, that they'll do them. A commitment nobody said out loud isn't a commitment.
Get the trade foremen in that room. Their input on what's really ready — and their read on the crew ahead of them — is worth more than anything you'll infer from a screen. Trade-level completion rates also tell you which sub is dragging the whole sequence, which is a conversation worth having early, not at closeout.
Why Software Matters Here (and Where It Doesn't)
None of this requires software. Plenty of great jobs run this loop with sticky notes on a wall and a whiteboard PPC chart, and if that's working for you, don't let anyone talk you out of it. What good look-ahead scheduling software buys you is the elimination of the transcription tax — the hours spent re-drawing the look-ahead every week by hand, and the errors that creep in when you do.
The features that actually serve the loop are specific: incomplete activities that roll forward automatically instead of being hand-copied; variance reasons captured at the moment you mark something incomplete, so they're never lost; PPC and variance trends that build themselves into a chart over the life of the job; and one shared view where the look-ahead and the weekly plan are genuinely the same underlying data, not two exports that drift apart. That last one is the whole game. A tool like LookAheadWall keeps the weekly plan and the rolling look-ahead as one connected model — and its trade-flow sequencing makes those "prerequisite not complete" handoff problems visible before they cost you a week — but the discipline has to come from you. Software that captures variance you never analyze is just a more expensive binder.
What Breaks the Loop
After twenty years watching this work and watching it fail, the failure modes are always the same three:
- Two disconnected systems. The look-ahead lives in one place, the weekly plan in another, and reconciling them is a manual chore nobody owns. They drift, then diverge, then one gets abandoned.
- Late updates. The look-ahead doesn't get updated before the next planning session, so you plan the new week off stale data. Everything downstream inherits that staleness.
- Variance captured but never analyzed. This is the quiet killer. You dutifully log reasons every week and never once total them up. You're collecting the medicine and never taking it.
The Bottom Line
Short-interval scheduling isn't a document you produce — it's a loop you run. This week's execution measures your plan. The reasons things missed diagnose your process. Those diagnoses reshape the look-ahead. And a cleaner look-ahead feeds a more reliable weekly plan, which produces better data, which sharpens the next cycle. Around and around, and every lap the whole job gets more predictable.
Give it eight or ten weeks of honest running and you'll watch your PPC climb, your handoff misses shrink, and your durations converge on reality. Not because the crews suddenly got better — because you finally closed the loop between what you planned and what actually happened, and let one teach the other.