A programme that sits in a drawer is worthless. Here's how to build a realistic construction programme and use it as a management tool, not a tender box-tick.
Every construction tender asks for a programme. Most contractors produce one. And on most small contractor projects, that's the last time anyone looks at it until the end-of-job argument about who caused the delay.
The construction programme should be the single most useful management tool on the project. It tells you what needs to happen, when, and in what order. It identifies the critical path. It highlights the pinch points. And when things go wrong — which they will — it shows you exactly where the impact falls.
Instead, on too many projects, the programme is a colourful bar chart produced at tender stage, based on optimistic assumptions, never updated, and completely disconnected from what's actually happening on site.
The most common mistake is starting with the dates and working backwards. "The client wants us finished by 1 September, so strip-out starts on 1 April." That's not programming — that's calendar-filling.
Start with the logic. What are the activities? What's the sequence? What are the dependencies? The strip-out must finish before the M&E first fix can start. The M&E first fix must be complete before the ceiling grid goes in. The ceiling tiles can't go in until the services above the ceiling are tested. These are logical relationships, and they drive the programme regardless of what date the client wants.
Once you have the logic right, add the durations. How long does the strip-out actually take with the labour you're planning? Not how long you'd like it to take — how long it takes based on experience and realistic productivity rates. If you've done similar projects before, use those durations as your reference. If you haven't, ask your subcontractors for realistic timescales and add contingency for the interfaces.
Only then do you put dates on it. If the logic and durations don't fit the contract period, you have a conversation about resources, sequencing, or completion dates. You don't just squash the bars to fit the calendar.
There's a sweet spot between a programme with 20 activities (too vague to manage anything) and one with 500 activities (too complex for anyone to maintain on a busy project).
For a typical commercial fit-out running 16 to 20 weeks, you want somewhere between 80 and 150 activities. Enough to capture every significant work package, broken down by location or floor where the work progresses through the building.
Here's a rough structure:
Pre-construction activities. Lead times for materials, design submissions, approvals. These get forgotten in programmes all the time, then become the reason the job starts late.
Enabling and strip-out works. Broken down by area if the building isn't being stripped all at once.
Structural and first fix works. By trade and by area. Partitions, M&E first fix, structural alterations — each as separate activity lines, showing the logical sequence as work progresses floor by floor or zone by zone.
Second fix and finishes. Again by trade and area. This is where the programme gets congested — multiple trades in the same spaces at the same time. Show the sequence clearly.
Testing and commissioning. Builders work, M&E commissioning, fire alarm, security, IT — each needs time, and they often need sequencing.
Handover. Cleaning, snagging, documentation assembly, PC inspection. Build this into the programme from the start, not as a two-day afterthought at the end.
The critical path is the longest sequence of dependent activities from start to finish. Delay any activity on the critical path and you delay the whole project.
On most fit-out projects, the critical path runs through either the M&E installation or the longest lead-time material delivery. Knowing which activities are critical changes how you manage the project — critical activities get the resource priority, the closest monitoring, and the earliest intervention when something slips.
Activities with float — spare time before they affect the critical path — are important too. But float gets consumed quickly. If the partitioning has two weeks of float and the subcontractor turns up a week late, you've already used half your buffer.
A programme that isn't updated is fiction. Weekly updates — showing actual progress against planned progress — are the minimum. This means:
Marking activities as started, showing percentage complete, and recording actual completion dates. When activities slip, updating the remaining logic and seeing where the impact falls. When new work is instructed, adding it to the programme and assessing the time effect.
The programme update should drive the weekly progress meeting. Instead of vague discussions about what's happening next week, you have a factual record: we planned to complete the partitions to level 3 this week, we achieved 80%, the remaining work carries over, the ceiling installation to level 3 slips by two days, but there's float so it doesn't affect the critical path.
That's useful information. "We're roughly on programme" is not.
Optimistic durations. The partitioning subcontractor says four weeks, you programme three because that's what fits. Congratulations — you've built a delay into the programme before the job starts.
No lead times. Specialist materials — raised access flooring, bespoke joinery, feature lighting — have lead times of 6 to 12 weeks. If these aren't in the programme, nobody's tracking them, and they'll be the reason you can't finish on time.
Ignoring the interfaces. The programme shows each trade as an independent bar. In reality, the M&E contractor needs access that the partitioning contractor hasn't finished clearing. The decorator needs the ceiling to be complete but the fire stopping hasn't been signed off. Programme interfaces — where one trade needs another to have finished something specific — are where delays breed.
No float. A programme with every activity back-to-back from start to finish has zero resilience. The first problem — a late material delivery, a week of bad weather, an RFI that takes too long — and the completion date is gone. Build in realistic float, particularly around high-risk activities and known pinch points.
Not sharing it. The programme only works as a management tool if everyone has access to it. Your site team, your subcontractors, the design team, the client — they all need to see the latest version and understand what's expected of them in the coming weeks.
You don't need Primavera P6. For most small contractor projects, a well-structured programme in Gantt chart format is perfectly adequate. The construction programme software matters less than the discipline of maintaining it.
What you do need is something that handles dependencies properly (so changes cascade through the logic), shows the critical path, and is easy to update weekly without a planning engineer.
In Construction AI, the project management module includes programme management linked to the rest of the project — so programme activities can reference drawings, RFIs, and subcontractor packages. When an RFI is overdue and it's linked to a programme activity, you can see the potential impact immediately. Combined with AI-powered insights, the system can flag programme risks before they become extension of time claims.
But whatever tool you use, the programme only works if someone owns it, updates it weekly, and uses it to manage the project. A realistic, maintained programme is worth more than any other single document on the job.
Stephen Mckenna MCIOB
30+ years in UK commercial construction, from site management to director level. Now building the project management tools he wished he'd had.
Tier-one contractor processes shouldn't only be available to tier-one contractors. Drawing registers, document control, RFIs, programmes, financial tracking — built for how you actually run projects, priced so any construction business can access them. Get started today.
By subscribing, you agree to our Privacy Policy.
The wrong drawing revision cost a six-figure rework bill. Here's how to set up construction document management — from naming conventions to digital control.
A submittal stuck in an architect's inbox for three weeks is a programme delay. Here's how to manage submittals so approvals don't hold up your project.
A bad RFI sits unanswered for a fortnight. A good one gets answered in days. Here's how to write RFIs that get responses and manage the ones that don't.
Tier-one processes, priced for any size business. Get started today.