Residential construction scheduling,
the complete reference.

A residential construction schedule, usually called the programme, is the job's model of time. It holds the tasks the build consists of, how long each realistically takes, the dependencies that say which tasks wait on which, and the milestones the rest of the job keys off, claim stages, inspections, selection deadlines, handover. Every date on the job is either computed by that model or guessed without it.

Defined precisely, a programme is not the chart. The chart is a rendering; the model is the tasks, durations and dependencies underneath it, and the difference shows the first time a date moves. A chart has to be re-drawn by hand; a model resequences itself and stays believable. The applied, step-by-step version of this reference lives in the residential construction scheduling guide; this page covers the discipline underneath it and the cluster of articles that carry the detail.

Why it matters

Time is the axis everything else on a residential job is plotted against. Materials are ordered against dates, trades book their weeks against dates, inspections are booked against dates, claims fall due at dated stages, and the costs of simply running the site accrue for every week the job stays open. A build that drifts a month is not late in the abstract; it is a month of supervision and site costs, a month of delayed claims, and a month of every downstream trade replanning around it. The schedule is also the only place all of those commitments meet, which is why a job without a credible one is run by phone call and memory instead.

The programme borrows its inputs from the front of the job. The scope comes from the contract, durations come from the business's own history of how long things really take, and the earliest believable dates come from supplier lead times. Its outputs run the back of the job. Order dates fall out of the schedule, worked backwards from when each material is needed on site, which is the whole subject of aligning procurement with the schedule. The trades' call-ups, the inspection bookings and the client's selection deadlines are all schedule outputs too.

The money side keys off the same dates. Time-related preliminaries accrue for every week the job runs, so schedule drift is a direct cost before any trade goes over budget. Progress claims fall due at stage milestones, so the timing of money in is the timing of the schedule, and any cash flow forecast is only as honest as the schedule dates underneath it. Even variations are half a scheduling subject, because most of them change time as well as scope, and a variation priced without its time cost is underpriced.

Most residential schedules are built once, for the contract, the bank or the client presentation, and never trusted again. The pattern is mechanical. The chart is drawn without real dependencies, the first delay makes it wrong, correcting it means dragging every downstream bar by hand, so it does not get corrected, and within a fortnight everyone on the job knows the dates are decorative. From then on the schedule exists for meetings and the job is actually run by phone call. Nobody decided to abandon it; it just stopped being worth reading.

The second death is quieter. The schedule stays roughly current, but only in the supervisor's head, updated by memory and reconciled at night, which is the same admin shift that eats the rest of a builder's evenings. A programme that lives in the office, or in one person's recall, runs nothing, because trades act on what they are told, not on what the office knows. The failure is the same one that lets a drawing revision sit in an inbox while the site builds from the superseded set, a missing path from known to communicated, and the drawing revision control guide covers that mechanism in its document form. Dates need the same protocol drawings do.

The third death happens before the job starts. In practice commencement is routinely held by open variations, unsigned change orders and final pricing that legitimately waits on the construction certificate, while the client, who expected everything locked before a start date, grows anxious and asks for the total. A signed but unstarted job is intent, not delivery, the same distinction the approvals versus completions guide draws at industry scale, and a business that measures its pipeline instead of its real progress carries the same blind spot inside its own walls. The pre-construction stall belongs on the programme, with owners and dates, exactly like the build itself.

The operators who schedule well hold an observation the chart never shows. The first delay is cheap; it is the response to it that is expensive or free, depending entirely on when it is known. A truss plant slipping five days costs almost nothing on the Monday it is known, because the carpenter can be redirected, the plumber moved before he loads the ute and the inspection rebooked before the slot is lost. The identical five days, discovered when the carpenter arrives to an empty slab, costs call-out fees, goodwill and often a queue position with every trade that follows. So they measure a schedule's value by one thing, how quickly a slip on site becomes a change in what every affected trade, supplier and the client believes happens next. Not how quickly the chart is corrected. How quickly the beliefs are.

Their working habits follow from that. The programme is updated from site facts on a short cadence, usually a weekly look-ahead per job covering the next two to three weeks, and the update ends with communication, not with a saved file. Causes of delay are recorded the day they happen, partly because that is what an extension of time claim later stands on, and partly because a schedule that only records effects teaches the business nothing. And they protect the baseline, because a plan edited in place until it agrees with reality is how a job reads as on time every single week and still finishes three months late.

Where software fits the workflow

Traditionally the programme is a wall chart or a spreadsheet, and every slip means an evening of re-drawing followed by a ring-around that misses someone. In VIABUILD the same loop runs connected. Construction scheduling holds the tasks with real dependencies, milestones and a preserved baseline, so moving one date resequences the chain and shows the drift honestly, and progress recorded on site through ViaSite updates the same programme the office reads, so the site and the schedule stop being two versions of the truth. The builder still decides every resequence; what disappears is the re-drawing and the ring-around.

Scheduling itself is not legislated, but several of its inputs are, and the wider delivery environment has been moving against builders for years. The points below are labelled by evidence class. Figures are point-in-time and the rules differ by state, so confirm current sources before relying on any of them.

• Legislation. Building work in each state and territory is subject to mandatory inspections at prescribed stages, and work generally may not proceed past a hold point until the inspection or certification has occurred. Which stages, and who inspects, varies by jurisdiction, so the hold points on a programme have to be built from the rules where the job is, and inspection bookings scheduled like tasks with lead times of their own. Confirm the prescribed stages with your state or territory building regulator.
• Legislation. Domestic building contract legislation in most jurisdictions requires the contract to state when work will be completed, or the building period, and standard residential contracts provide for claiming extensions of time for causes beyond the builder's control, usually with notice requirements and timeframes. A programme that records causes of delay as they happen is the evidence those claims stand on; one updated by memory is not. Confirm the requirements in your jurisdiction and contract.
• Government statistics. Analysis of ABS building activity data indicates the average build time for a detached house rose by roughly 50 per cent between September 2019 and June 2024. The figure is point-in-time and methodology-dependent; confirm against current ABS building activity data before quoting it. Whatever the precise number, the direction is the operating reality residential programmes are now built in, longer lead times and slower delivery than the durations many builders still copy from earlier years.
• Government policy. The National Housing Accord targets 1.2 million new homes over five years from July 2024, and reported completions have tracked below the pace that target implies. The figures are point-in-time; confirm against current government reporting. The relevant lesson for a single business is the one the approvals versus completions distinction teaches, an approval is permission, not a delivered home, and only commencements and completions describe delivery.
• Common practice. The Australian residential industry effectively shuts down over the Christmas and January break, and northern jobs carry wet season constraints. Programmes that treat the calendar as fifty-two even weeks slip by design, which is why experienced schedulers plan stage targets around the shutdown rather than discovering it in December.

Illustrative only, not a benchmark. A home is a week from frame stage when the truss plant advises a five working day slip on delivery. In the first version the supervisor hears on Monday and the programme carries real dependencies, so the effect is computed in minutes. The frame carpenter is moved to another of the builder's jobs for the week, the plumber's rough-in is shifted before he plans his run, the frame inspection is rebooked before the slot lapses, and the client is told the frame claim will move a week, before they notice a quiet site and start wondering. The delay costs five days and almost nothing else.

In the second version the plant's email sits unread and the schedule is a chart from the contract. The carpenter arrives Monday to an empty slab and charges the call-out, or spends the goodwill that would have absorbed the next favour. The plumber turns up a fortnight later to a frame that is not ready because nobody moved him. The inspection booking lapses and the next available slot adds four days to the five. The client hears about the delay last, from the site, and now reads every future date with suspicion. Same supplier, same five days. What differed was when the slip was known and how quickly it changed what everyone downstream believed happens next, and that difference, repeated across every delay on every job, is most of what separates builders who deliver near their dates from builders who explain them.

Programme (the schedule, the job's model of time), task (a unit of scheduled work), duration (how long a task takes), dependency (the rule saying a task waits on another), critical path (the chain of dependent tasks that sets the end date), float (how far a task can slip without moving the end date), milestone (a date the rest of the job keys off), hold point (a gate work may not proceed past until inspected or certified), lead time (the gap between ordering something and having it on site), baseline (the preserved original plan), drift (the measured gap between baseline and live dates), extension of time (a contractual claim to move the completion date). The wider vocabulary lives in the construction glossary.

The natural next article is the critical path and task dependencies, the logic that decides which of the dates on this page actually matter.

• Australian Bureau of Statistics , building activity data, the source series for dwelling approvals, commencements, completions and build times.
• Australian Government Treasury , the National Housing Accord and reporting against its target.
• Your state or territory's building regulator, for the mandatory inspection stages, certification hold points and domestic building contract time requirements that apply where the job is built.