Real scheduling is not a Gantt chart someone updates when they remember to. It is one baseline, tracked against reality every week, with the statutory dates that actually matter built into it.
Key takeaways
STEP 01 OF 10
A scheduling function needs a named owner with the authority to push back on an unrealistic sequence, not a task quietly absorbed into a PM's existing workload with no dedicated time. On a smaller firm this can be a part-time responsibility, but it needs to be explicitly assigned, the same way document control needs a named owner in a document control system for a GC.
The test for whether scheduling is genuinely a function rather than an afterthought: does anyone have the authority to say a proposed sequence will not work, before the crew is committed to it, or only after it has already failed on site?
STEP 02 OF 10
Procore's Schedule tool documentation confirms it imports existing schedules created in Microsoft Project, Primavera P6, and the MPX file format — the critical-path logic itself is built in one of those dedicated tools, not invented in a general project-management app. A firm standing up scheduling for the first time should budget for one of these tools directly, and treat any project-management software's own schedule view as a display layer for that imported logic, not a replacement for it.
This distinction matters because a spreadsheet-built "schedule" with dates typed in by hand has no actual critical-path calculation behind it — it looks like a schedule and cannot tell you which delay actually moves the completion date.
STEP 03 OF 10
Only one Baseline Start is supported natively in Procore's Schedule tool — a genuinely useful constraint to design around, not a limitation to work around. Set the baseline once, at the point the schedule is first approved for the project, and treat any later comparison as "current versus baseline," not a moving target that gets reset every time the schedule slips.
A schedule that gets re-baselined every time it falls behind is not tracking performance, it is erasing the record of it. Keep the original baseline fixed and measure variance against it for the life of the project.
STEP 04 OF 10
Every provincial lien, holdback and closeout deadline in this cluster runs from a substantial-performance or completion date — see project closeout in thirty days for the full downstream calendar this date drives. A schedule that treats this as an implicit endpoint rather than an explicit, tracked milestone gives the whole statutory calendar nothing concrete to run from.
Update this date the moment the schedule itself moves, not at the next scheduled status meeting. A stale substantial-performance date on the schedule while the actual completion date has already shifted is exactly how a lien-deadline calendar drifts out of sync with reality.
STEP 05 OF 10
Alberta's statutory substantial-performance test is a formula against the contract price — 3% of the first $500,000, 2% of the next $500,000, and 1% of the balance. See setting up a change management process for how this threshold moves as approved change orders raise the contract sum. A scheduler tracking this date on an Alberta project needs the current cumulative contract price feeding into the calculation, not the number from the original bid.
This is a genuinely different discipline than Ontario, where the Construction Act has no equivalent percentage test at all — a scheduler working across both provinces should not build one universal formula into the schedule template.
STEP 06 OF 10
A required inspection or quality sign-off — see building a quality control programme — should appear as a scheduled activity with its own duration and dependency, not an assumption that it happens automatically between two other tasks. A QC element that consistently scores low because the schedule never actually allows time for it is a scheduling failure wearing a quality label.
This is where scheduling and quality control genuinely meet: a schedule that squeezes inspection time to hit a milestone is setting the quality programme up to either fail the inspection or skip it, and neither outcome is one the schedule should be quietly forcing.
STEP 07 OF 10
Delay entries and productivity entries from the daily reporting system are a direct input into how the schedule gets updated — see rolling out digital daily reporting for the field-level record this depends on. A scheduler updating the critical path from a verbal Friday-afternoon summary is working from a much thinner record than one pulling structured delay data logged the same day it occurred.
This connection is also what makes a delay claim defensible if one ever arises — a dated, field-logged delay entry tied to a specific schedule activity is a fundamentally stronger record than a reconstructed narrative built after the fact.
STEP 08 OF 10
A weekly schedule update — actual progress against the baseline, current critical path, upcoming hold points — catches a slipping activity while there is still time to recover float. A monthly or ad hoc review tends to discover the slip only once it has already consumed the float that would have absorbed it.
Keep this review short and structured: what moved, what is now on the critical path that was not last week, and what decision, if any, is needed this week. A scheduling review that turns into a general project meeting stops being a scheduling review.
STEP 09 OF 10
A schedule that only reports dates hides the information that actually matters for decision-making: how much slack exists before a delay on a specific activity threatens the completion date. Report float alongside dates in every update, particularly for activities feeding into the substantial-performance milestone from step four.
A scheduling function that reports float explicitly is one that can tell a PM, honestly, whether a two-day delay on a specific trade is a non-event or a genuine threat to the completion date — a distinction a dates-only schedule cannot make.
STEP 10 OF 10
At project closeout, reconcile the baseline against the as-built schedule and archive both, alongside the substantial-performance date that ultimately governed the province's statutory deadlines. See project closeout in thirty days for how this feeds the broader closeout sequence.
This reconciliation is also the record a firm should be reviewing across projects to improve its own estimating and sequencing over time — a scheduling function that never looks back at how its baselines compared to reality is not actually learning from the projects it runs.
Re-baselining the schedule every time it falls behind. A moving baseline erases the record of variance instead of measuring it — set it once and track against it for the life of the project.
Building schedule logic in a spreadsheet instead of importing from a dedicated CPM tool. A hand-typed date list has no actual critical-path calculation behind it and cannot say which delay genuinely threatens completion.
Leaving the substantial-performance date implicit rather than an explicit, tracked milestone. Every provincial lien and holdback deadline in this cluster runs from that date — a schedule that never names it gives the statutory calendar nothing to run from.
Using Alberta's substantial-performance percentage formula on an Ontario project, or vice versa. Ontario's Construction Act has no equivalent percentage test — carrying Alberta's formula across the border misstates the threshold entirely.
Reporting only dates and never float. A dates-only schedule cannot distinguish a two-day delay that is a non-event from one that genuinely threatens the completion date.
Step five asks a scheduler to recalculate Alberta's substantial-performance threshold whenever the contract price moves — here is that recalculation in full.
A schedule is built against an original $1,200,000 contract, with a substantial-performance threshold of $27,000 under Alberta's tiered formula — 3% of the first $500,000, 2% of the next $500,000, 1% of the remaining $200,000. Midway through the project, an approved $150,000 change order raises the contract price to $1,350,000. Re-run against the new price, the threshold becomes $28,500 — the same tiers applied to a larger base. If the scheduler does not update this figure, the schedule is tracking substantial performance against a threshold that is now $1,500 out of date, which can shift exactly when the milestone is legitimately reached.
This is a small dollar shift in isolation, but the mechanism is the point: any approved change order on an Alberta project is a scheduling input, not just a budget line, because it moves a statutory milestone the whole downstream lien and holdback calendar depends on.
A scheduling template built for one province and reused nationally will carry the wrong milestone logic into projects where it does not apply.
The scheduling software and the baseline discipline can be identical across every province a firm works in — the milestone logic feeding the statutory calendar cannot.
The full baseline-and-float discipline scales down for a short project, but naming the substantial-performance or completion date explicitly still matters — it is what the lien deadline runs from regardless of project size.
A PM can hold the role part-time, provided the authority to push back on an unrealistic sequence comes with it — the title matters less than whether the function has real decision-making weight.
Rarely, and only for a genuinely fundamental scope change that makes the original baseline meaningless for comparison — a schedule slipping from normal field conditions is not, on its own, a reason to reset it.
Based on its own documentation, it imports schedules built in those tools rather than replacing their scheduling engine — treat it as the display and coordination layer, not the source of the critical-path calculation itself.
A 30-minute call is enough to tell you whether it is worth building.
