Construction project management software, scoped to the contract

Ask a contractor what software they need and you will hear about site progress. Photographs, daily reports, a programme that is already out of date. All of that is real, and none of it is why these projects get commissioned. They get commissioned because somebody cannot answer, on demand, how much of this job we have earned, how much we have been paid, and what we are owed that has not been certified yet.

That is a contract accounting problem wearing a site management costume. Scope it the other way round and the build goes badly.

Why the money side comes first

The UK Insolvency Service publishes company insolvencies by industry, and construction has been at the top of that table for a long time. In the twelve months to May 2026 the sector recorded 3,803 insolvencies, 17 percent of all cases where the industry was captured, ahead of wholesale and retail trade at 3,527 and accommodation and food service at 3,296.

For context, the overall company insolvency rate in that period was 50.9 per 10,000 companies, or one in 196, down from 53.0 a year earlier. So the construction number is not a story about a bad year across the economy. It is a sector that sits at the top of the table while the general rate falls.

Contractors do not usually fail because they cannot build. They fail because of timing: work certified late, a variation argued about for two months, retention released a year after practical completion, all while subcontractors and suppliers need paying now. Software cannot fix a thin margin, but it can stop a firm from discovering a cash gap three weeks after it opened.

The payment clock is a legal object

This is the part most generic project tools get wrong. Payment dates in construction are not a preference you configure, they are statutory in some jurisdictions, and they run on notices rather than invoices.

Part II of the UK Housing Grants, Construction and Regeneration Act 1996 is the clearest written example. Section 109 gives a party the entitlement to payment by instalments, stage payments or other periodic payments, unless the work is specified or agreed to last under 45 days. Section 110 requires every construction contract to provide an adequate mechanism for determining what payments become due and when, and to "provide for a final date for payment in relation to any sum which becomes due".

Then the notice machinery. Section 110A requires a payment notice not later than five days after the payment due date, stating the sum considered due and the basis on which it is calculated. Section 110B lets the payee issue its own notice if the payer does not, and postpones the final date for payment by the number of days that notice was late. Section 111 obliges the payer to pay the notified sum on or before the final date for payment unless it has served a notice of intention to pay less, which must itself state the sum considered due and the basis for it.

Two consequences follow, and both are software requirements rather than legal trivia. A missed notice has a financial value, because the notified sum becomes payable. And a date that moves because a notice was late is a computation, not a field somebody types in.

What the escalation path implies

Section 112 lets an unpaid payee suspend performance after giving at least seven days notice stating the grounds, with the payer liable for the reasonable costs of the suspension. Section 113 makes a clause conditioning payment on the payer receiving payment from a third person ineffective, except where that third person is insolvent, which is the provision that killed pay-when-paid.

And section 108 lets either party refer a dispute to adjudication at any time: the adjudicator appointed and the dispute referred within 7 days of the notice, a decision within 28 days of referral, extendable by up to 14 days with the referring party's consent, binding until finally determined by litigation, arbitration or agreement.

A system that cannot produce the notice history for a single valuation on demand is a system that cannot support an adjudication, and 28 days is not long enough to reconstruct it from a shared drive.

Variations are where the money leaks

Every contractor has the same story. An instruction given verbally on site, work done in good faith, and a dispute six months later about whether it was a variation or always in scope. The software job here is narrow and high value: make the instruction capturable at the moment it happens, with who gave it, when, against which contract clause, and what it is expected to cost.

It also matters that variation entitlement is often capped. Saudi Arabia's procurement reforms are a live example: the maximum contract variation is reported to be moving from 10 percent to 20 percent under the new Government Tenders and Procurement Law. If your contracts carry a cap, the system has to track cumulative variation value against it, because the figure that matters is the running total, not the individual instruction.

What belongs in phase one

The first five are a contract register and a few derived views over it. That is a small system. The temptation is to start at the bottom of that table because site photographs are the visible pain, and it is the wrong order: progress capture without the money model produces a prettier version of what the site already has.

Retention, the money you cannot see

Retention is a percentage withheld from each payment, released in part at practical completion and the balance at the end of the defects period. On a single job it is a rounding error. Across a dozen jobs over three years it is frequently the largest asset the firm never looks at, because it is spread across valuations in a spreadsheet nobody reconciles.

Model it as its own ledger with a due date per tranche, and set a reminder before each release date rather than after. This is the single highest return item on the list and it is almost always deferred, because unlike a variation dispute nobody is shouting about it.

Where the real cost sits

Not in the screens. In the integration with the accounting package, and in the data migration from however many spreadsheets currently hold the truth. The ranges and what moves them are in what an integration between two systems really costs, and the invoice side specifically tends to be fiddly, which is covered in automating invoice generation and the three places it breaks.

The migration deserves its own line in the budget. Contractors keep years of valuation history in workbooks with formulas nobody can explain, and that history is needed for the retention ledger to be correct on day one. The approach that survives is in migrating from spreadsheets without losing the history. Before any of it, write the scope down: how to scope a software project before anyone writes code is the step that most often gets skipped here, because the requirement feels obvious and is not.

The commercial terms you are modelling are not a software question at all, and getting them wrong upstream makes the build pointless. BDG Labs set out the regional reality of that clock in what payment terms for B2B clients in the Gulf really mean, which is worth reading before you encode any default into a contract register.

An honest limit

Most contractors should not build this. If you run fewer than about ten concurrent jobs, a well structured contract register in a spreadsheet plus calendar reminders for notice dates and retention releases will get you most of the benefit for none of the cost. The decision order is in custom software versus off the shelf, decided in the right order, and the honest answer for a small firm is usually off the shelf or nothing yet.

The case for building is specific: you work under several contract forms with genuinely different payment mechanics, your retention exposure across jobs has become material, or you have an accounting system no construction product integrates with. Short of that, the reasons projects like this fail are the ordinary ones, set out in why software projects fail, and why the average is the wrong number.

One jurisdiction note, because it would be misleading to leave it out. The Act quoted above is UK legislation and does not apply in Egypt, the Gulf or anywhere else by default. Contracts in those markets may adopt similar mechanics by agreement, or FIDIC forms with their own notice regimes, or nothing comparable at all. Read it as a model of how precisely a payment clock can be specified, then build against the contracts you actually hold. The primary text is Part II of the Housing Grants, Construction and Regeneration Act 1996, and the insolvency figures are in the Insolvency Service's commentary on company insolvency statistics for May 2026.