Every look-ahead process eventually produces a number. The question is whether that number tells you something you can act on, or whether it's just a report you print, staple, and forget on the trailer table. I've sat in plenty of production meetings where the schedule metric on the wall was a green bar that meant nothing — the crew was three days behind, everyone in the room knew it, and the software still said 92%. That's the trap. The metrics below are the ones that actually move a job. The rest is decoration.
Percent Plan Complete: the one number to start with
If you track exactly one thing, track Percent Plan Complete. PPC is completed commitments divided by total commitments for the week — and the word commitments is doing all the work. It is not "activities on the schedule." It is the specific tasks a foreman looked you in the eye on Monday and said, "Yes, we'll get that done this week." If your electrician committed to ten items and finished eight, your PPC for that trade is 80%. Simple to calculate, brutal to fake once you understand it.
Here's the part people miss: PPC is binary and it's unforgiving on purpose. A task that's 90% done counts as zero. You either made the commitment or you didn't. That feels harsh the first time a foreman gets dinged for a task that was "basically finished," but it's the whole point — a wall that's 90% painted still can't have carpet run in front of it, and the tile guy who's 90% ready still can't start. Partial credit is how schedules quietly slip a week at a time while everyone reports progress.
For where you should land: crews new to short-interval planning routinely start around 50-60%. That's normal and it's not a sign the process is broken — it's a sign the process is finally telling you the truth. Sustained work gets you to 80%, and genuinely disciplined teams live above 85%. But don't chase the number by itself. A PPC that jumps from 55% to 90% in three weeks usually means someone learned to stop committing to anything risky, which brings us to the failure mode that ruins this metric.
The sandbagging problem
The fastest way to a beautiful PPC is to commit to almost nothing. If a foreman only promises the four tasks he's dead certain of and stays quiet about the other six, he'll hit 100% and your job will still fall behind because the real work isn't getting planned. Watch PPC and total commitment volume together. If PPC climbs while the number of weekly commitments quietly drops, you're not getting better — you're getting sandbagged. Call it out gently and early. The fix is culture: make it clear that a missed commitment with a good reason is useful data, not a black mark.
PPC by trade — where coordination problems hide
A single project-wide PPC hides more than it reveals. Break it out by trade and the story sharpens fast. When your framers sit at 88% and your mechanical crew is stuck at 55%, you don't have a "planning problem" — you have a mechanical problem, and now you know exactly where to spend your Monday.
Better yet, watch the handoffs between trades. Some of the ugliest recurring misses aren't inside one trade — they're at the seam where one trade flows into the next. Rough plumbing that never quite clears its inspection on time so the framer can't close the wall. Fireproofing that runs long so the drywall crew keeps getting bumped. When you map your work as connected trade flows rather than a flat list, these dependency failures stop being surprises and start showing up as a predictable pattern you can attack. That's the whole reason to sequence trades visually in the first place — you see the collision before it happens, not in the variance report afterward.
Variance reasons: the metric that actually teaches you
PPC tells you that you missed. Variance reasons tell you why, and the "why" is where all the improvement lives. Every time a committed task doesn't complete, capture the reason before the meeting ends — memories are already fuzzy by Thursday and gone by the next Monday. Keep the categories short and consistent so you can actually count them:
- Prerequisite work — the predecessor trade wasn't finished (the most common one, by far)
- Materials — wrong, late, damaged, or not on site
- Manpower — crew didn't show, got pulled to another job, or was short-handed
- Information/RFI — waiting on a detail, submittal, or answer from the design team
- Equipment — lift, hoist, or tool wasn't available
- Directive change — the owner or a change order moved the work
- Weather — the one nobody controls
The single reason won't teach you much. The pattern over a month is gold. If 40% of your misses trace back to prerequisite work, your problem isn't the trade that missed — it's the trade upstream, and no amount of leaning on the downstream foreman will fix it. If a third of your variance is materials, your problem is procurement and lead times, and it needs solving in the office, not the field. This is the difference between a team that argues about the same delay every week and a team that actually removes it. Chase the category, not the individual task.
Constraint resolution timing — the leading indicator
PPC and variance are lagging indicators. They tell you how last week went. Constraint resolution timing is one of the few things you can measure that tells you how next week is shaping up, which makes it worth more than its reputation.
The make-ready process exists to clear obstacles before work is scheduled — materials confirmed on site, prior trade complete, inspection passed, submittal approved, equipment reserved. The metric that matters is how many weeks of lead time you have when each constraint gets cleared. A constraint resolved three weeks out is a non-event. The same constraint resolved the morning the crew shows up is a fire drill that probably cost you the day anyway.
Track two things and you'll know if your look-ahead is doing its job. First, the late-resolution rate: how often constraints get cleared in the same week the work is supposed to start — too late for anyone to plan around comfortably. Second, and this is the one that hurts, how many tasks enter the commitment week with a constraint still open. That number should be near zero. A rolling look-ahead that keeps letting constrained work into the current week isn't a plan — it's a wish list. If you're seeing that regularly, your make-ready is happening too shallow or too late, and no software feature fixes a make-ready conversation that nobody's having.
What not to measure
Vanity metrics are seductive because they always go up and to the right, and they never tell you to do anything. Ignore them:
- Number of activities in the look-ahead. A schedule with 400 line items isn't more planned than one with 120 — it's usually just harder to read. Density is not diligence.
- Number of constraints tracked. Logging a constraint that was never going to stop the work is busywork dressed up as rigor. Track the ones that would actually halt a crew.
- Reports generated. Nobody ever hung drywall faster because the trailer printed a nicer report. A report only counts if it changed a decision.
- Login counts and screen time. Adoption matters, but time-in-app measures the tool, not the job.
The tell for a vanity metric is simple: ask "if this number doubled, would we do anything differently?" If the answer is no, stop measuring it.
Trends beat snapshots
Any single week's PPC is noise. A framing crew hits a bad-weather stretch, a big material drop slips, and the number tanks for reasons that have nothing to do with planning quality. What you're really after is the slope. Is PPC climbing month over month? Are the same three variance reasons shrinking as you solve them, replaced by new and smaller ones? Is the gap between your estimated durations and your actual durations narrowing as your foremen get better at calling their own work?
That last one — duration estimation accuracy — is a quiet marker of a maturing team. Early on, everyone's guesses are optimistic. As people watch their own predictions bump against reality week after week, the guesses tighten. When your planned durations start matching actuals within a reasonable margin, your whole schedule gets more trustworthy, and every commitment downstream firms up with it. Software that keeps this history automatically — planned versus actual, week over week — turns a fuzzy gut feeling into a line on a chart you can show the crew. LookAheadWall keeps that record so the trend is there when you need it, but the discipline is yours: the tool measures, the team improves.
Making the numbers actually run the job
A metric only earns its keep if it drives a decision, so build the review to force that. A few rules of thumb that hold up on real jobs:
- Capture automatically, review deliberately. If your foremen are hand-tallying PPC in a spreadsheet Sunday night, the data will be late, wrong, and resented. Let the system compute it from the commitments; spend the human time on interpreting it, not producing it.
- Review operational metrics weekly, trends monthly. PPC and variance belong in every weekly work-plan meeting. Slopes and estimation accuracy are a monthly conversation — looking at trend lines every week just amplifies the noise.
- Put every metric in front of someone who can act on it. A foreman needs his own trade's PPC and his own open constraints. The owner needs milestone adherence and the trend. Sending everyone everything guarantees nobody reads any of it.
- End every review with a verb. The meeting isn't over when you've read the numbers — it's over when someone owns an action because of them. "Materials variance is up, so I'm calling the supplier and moving next week's slab pour behind the block delivery." That's a metric doing its job.
Set targets from your baseline, not from a slide
Before you set a goal, measure where you actually are for a few weeks. A crew averaging 55% PPC does not get to 85% by having someone announce that 85% is the new target — they get there through 60%, then 65%, then a plateau, then 70%. Set the next honest step, hit it, and set the one after. And keep watching for the gaming that quiet targets invite: the undercommitting we talked about, or reclassifying a real miss as "weather" to protect a number. The point of all this was never a good-looking dashboard. It's a job that finishes closer to on time because the plan was believable and the misses got fixed. Keep the metrics pointed at that, measure a few things well instead of many things poorly, and act on what they tell you — that's the whole game.