Gantt Chart for Construction Projects
A construction schedule is the one document every trade, supplier and inspector on the job reads the same way. Get it right and the concrete cures before the frame lands on it, the electrician arrives after the walls are open, and the flooring crew never trips over wet paint. Get it wrong and you pay in idle crews and rework. A Gantt chart is the tool that turns that sequence into dates everyone can trust, provided you build it to reflect how a site really works.
Why construction lives or dies by sequence
Almost no trade on a building site can start whenever it likes. The plumber needs open walls, the walls need a frame, the frame needs a cured slab, and the slab needs a prepared site. Construction is a chain of physical facts, and a schedule that ignores those facts is not a plan, it is a wish. A Gantt chart is the natural home for this kind of work because it shows two things at once: how long each activity takes, and what must finish before the next crew shows up.
The reason a bar chart beats a to-do list here is that a building has no shortcuts through physics. You cannot paint a wall before it exists, and you cannot pour a footing into an unexcavated trench. When the sequence is drawn as linked bars, a slip in one trade visibly pushes every trade downstream, which is exactly the warning a site manager needs before crews turn up to a job that is not ready for them.
The standard phases of a build
Most vertical construction projects move through the same broad phases, whatever the size of the building. The names vary by region and contract, but the order rarely does. A Gantt chart usually groups tasks under these phases as summary bars, then breaks each one into the detailed activities the trades actually book against.
| Phase | What happens | Typical duration | Waits on |
|---|---|---|---|
| Site preparation | Survey, clearing, excavation, temporary services | 1 to 3 weeks | Permits, site access |
| Foundation | Formwork, rebar, pour, cure, backfill | 2 to 4 weeks | Site prep complete |
| Structure | Frame, floors, roof deck, weathertight envelope | 4 to 10 weeks | Foundation cured |
| MEP rough-in | Mechanical, electrical, plumbing inside open walls | 3 to 6 weeks | Structure and roof |
| Finishes | Insulation, drywall, paint, flooring, fixtures | 4 to 8 weeks | MEP inspected |
| Handover | Commissioning, snag list, final inspection, keys | 1 to 2 weeks | Finishes complete |
These figures are starting points, not promises. Estimating each activity honestly is its own skill, and the numbers above only hold if you have checked them against your crew sizes and local conditions rather than copying them blind.
Dependencies are the backbone of the plan
In a construction Gantt chart, the links between tasks carry more meaning than the bars themselves. A finish-to-start link says the frame cannot begin until the foundation is cured. A start-to-start link with an offset says drywall can begin two days after the first room passes its electrical inspection, not all at once. Getting these links right is what makes the schedule recompute itself when reality shifts.
The most common mistake is drawing dependencies loosely, or worse, not at all, and pinning every task to a fixed calendar date instead. That produces a chart that looks tidy and lies quietly. When the excavation runs three days long, a properly linked plan slides the foundation, the frame and everything after it automatically. A plan of fixed dates just leaves stale numbers behind while crews arrive to unready work.
Lag time: curing, drying and inspections
Some of the most important time in a build is time when nobody is working. Concrete needs days to cure before it can bear load. Screed and plaster need to dry before the next layer goes on. Inspections sit between phases and cannot be rushed. On a Gantt chart these gaps are modelled as lag, a deliberate delay on a dependency rather than a task with a crew assigned to it.
Lag matters because it is invisible to anyone reading a simple bar chart. A footing pour might take one day, but the seven day cure that follows is what actually governs when the frame can start. If you draw the pour and the frame back to back with no lag, your schedule will be a week optimistic on the most load-bearing decision in the project. Model the wait explicitly and the plan tells the truth.
Long-lead procurement runs in parallel
Structural steel, curtain wall, switchgear, elevators and custom joinery can take twelve to twenty weeks or more from order to delivery. If you wait until the trade is due on site to place the order, you have already lost the job. Long-lead items belong on the Gantt chart as their own bars, starting near day one and running in parallel with the physical work, with a delivery milestone that a trade depends on.
The discipline is to link the delivery milestone to the activity that consumes the material, so a procurement slip shows up as a threat to the build date rather than a surprise in month four. A good schedule makes the buyer and the site manager read from the same page: the day steel is late is the day the frame is at risk, and everyone can see it coming.
Sequencing trades so they do not clash
Two crews in the same room is not twice the progress, it is a safety incident and a quality problem waiting to happen. The electrician and the plumber both want the walls open, but they cannot occupy the same wall cavity at the same moment, and the drywall crew must wait for both to finish and pass inspection. Sequencing is the art of packing trades tightly without overlapping them where they physically conflict.
The usual pattern is to phase the work by zone. Floor one gets its MEP rough-in while floor two is still framing, then the finishes crew follows the MEP crew up the building a floor behind. This keeps every crew busy without stacking them on top of each other. Modelling zones as parallel bars with staggered starts is what turns a cramped site into a smooth production line.
Building weather contingency into the schedule
Groundworks, concrete pours, roofing and external cladding are all exposed to weather. A schedule that assumes perfect conditions will slip in the first wet week and never recover. There are two honest ways to protect against this, and both belong on the chart rather than in the site manager's head.
- Add a weather contingency buffer as a visible task after each weather-exposed phase, sized from local seasonal data rather than optimism.
- Sequence weather-sensitive work into the drier months where the calendar allows, and push interior finishes into the wet season.
- Track buffer consumption at every update so you know how much protection remains before the completion date is genuinely at risk.
- Keep the buffer as its own bar, never buried inside task estimates, so nobody quietly spends it without anyone noticing.
A buffer you can see is a buffer you can manage. A buffer hidden inside padded durations is just a lie you tell yourself, and it disappears the first time a task is questioned.
A worked mini construction schedule
Here is a compressed schedule for a small single-storey extension, the kind of job where the sequence is easy to follow but the lag and procurement traps are still real. Every date flows from the links, not from a wish.
The plan. Six activities, linked finish-to-start except where noted, starting Monday March 2.
| Task | Duration | Depends on | Note |
|---|---|---|---|
| Excavate footings | 3 days | Start | Weather exposed |
| Pour foundation | 1 day | Excavate | Then 7 day cure lag |
| Erect frame and roof | 8 days | Pour + 7 lag | Steel ordered week 1 |
| MEP rough-in | 5 days | Frame | Walls open |
| Finishes | 7 days | MEP + inspection | Interior, weatherproof |
| Handover | 2 days | Finishes | Snag and sign-off |
The read. Excavation and the one day pour take four working days, but the seven day cure means the frame cannot start until roughly March 16, not March 6. That single lag governs the whole downstream chain. Order the steel in week one and it arrives comfortably before the frame is due.
What happens when one task slips
The value of a linked schedule is proven the moment something goes wrong, which on a building site is most weeks. Suppose the excavation in the extension above hits rock and runs two days long. In a chart of fixed dates, that fact sits quietly in one bar. In a properly linked Gantt chart, the delay ripples through every task that follows.
The slip. Excavate footings runs 5 days instead of 3, finishing two days late.
The cascade. The pour moves two days later, so the cure lag ends two days later, so the frame, MEP, finishes and handover all shift two days. The completion date moves from its original target to two days beyond it, and the chart shows it instantly.
The decision. Because the slip is visible, the site manager can act: pull finishes forward by adding a second crew, or spend two days of the weather buffer if one was set. Without the links, that two day slip would surface as a missed handover with no warning and no room to respond. This is the difference between a plan that thinks with you and one that merely records the past.
The critical path on a building site
Not every task controls the finish date. The critical path is the longest chain of dependent activities, and it is the only chain where a one day slip pushes the whole project one day later. On the extension, the path runs excavate, pour, cure, frame, MEP, finishes, handover. A delay to any of those moves completion. A delay to a side task with spare time, such as ordering door handles, does not.
Knowing the critical path tells a site manager where to spend attention and money. Adding a crew to a task with plenty of float buys nothing. Adding it to a critical task, such as speeding the frame with a larger gang, actually pulls the finish date in. Every acceleration decision on site should start with the question of whether the task is even on the critical path.
Keeping the schedule alive during the build
A construction Gantt chart is not a document you draw once and pin to the site hut wall. It is a live model that only earns its keep if it is updated. The best sites run a short weekly cycle: mark what actually finished, re-forecast what is left, and read what moved. A schedule updated every week catches a two week slip while there is still time to react. A schedule updated never catches it at handover.
Tracking actual progress against the original dependency plan is what turns the chart from a hopeful drawing into an early-warning system. You do not need heavy software to do it. You can build and maintain a fully linked construction schedule in a free browser-based Gantt maker, keeping the phases, lags and procurement bars all in one view that every trade can read.
Common construction scheduling mistakes
Most broken construction schedules fail for the same handful of reasons, and all of them are avoidable once you know the pattern. Watch for these before they cost you a crew day.
| Mistake | What it causes | Fix |
|---|---|---|
| No cure or drying lag | Frame or finishes start too early on paper | Model every wait as explicit lag |
| Fixed dates instead of links | Slips do not cascade, dates go stale | Link tasks and let the plan recompute |
| Procurement left off the chart | Trades arrive before materials do | Add long-lead bars with delivery milestones |
| Buffer hidden in task padding | Contingency spent invisibly, then gone | Show weather buffers as their own bars |
| Stacked trades in one zone | Safety and quality problems, rework | Phase work by zone with staggered starts |
None of these needs a bigger budget to fix. They need a schedule that reflects how a site actually behaves, honestly maintained, and read by everyone who books work against it.
Templates that use this
Keep reading
- Gantt Chart Dependencies Explained: FS, SS, FF and SF
- How to Estimate Task Durations (Without Guessing)
- Critical Path Method (CPM): Steps, Formula & Example
Frequently asked questions
Why do I need lag time on a construction Gantt chart?
Because much of the time in a build is spent waiting, not working. Concrete cures, plaster dries and inspections sit between phases. If you draw the pour and the frame back to back with no lag, the schedule is optimistic on your most load-bearing decision. Model each wait as explicit lag so the dates stay honest.
How do I stop two trades clashing in the same space?
Sequence the work by zone rather than running everything at once. Let the MEP crew work floor one while floor two is still framing, then send finishes up a floor behind. Model each zone as a parallel bar with a staggered start so crews stay busy without occupying the same space at the same moment.
Where should long-lead procurement go on the chart?
On its own bar, starting near day one and running in parallel with the physical work, ending in a delivery milestone. Link that milestone to the activity that uses the material. A procurement slip then shows up as a direct threat to the build date instead of a surprise weeks later when the crew arrives to no materials.
How much weather contingency should I add?
Size it from local seasonal data rather than optimism, and place it as a visible buffer bar after each weather-exposed phase such as groundworks or roofing. Track how much of the buffer you consume at every update. A buffer you can see is one you can manage; padding hidden inside task durations vanishes the first time an estimate is questioned.
Can I run a real construction schedule in a free tool?
Yes. A capable free Gantt maker handles linked dependencies, lag on those links, milestones for deliveries and summary bars for phases, which is everything a small to mid-size build needs. The discipline matters more than the price: draw honest dependencies, model your lags, and update the chart weekly so it stays a live plan rather than a drawing.