Signature is the halfway point, not the finish line.
Last updated
Zepth Vector module
9.2%
of annual revenue lost to ineffective contract management
World Commerce & Contracting
Top performers run around 3%; laggards 15–20%. On large capital projects the leakage reaches roughly 15%.
£3.2–5.9bn
held in retention across the UK construction industry at any one time
UK BEIS / Pye Tait retention research
With roughly £240m a year lost to upstream insolvency — money already earned, and never paid.
44%
of surveyed UK contractors had lost retention money to an upstream insolvency
UK BEIS / Pye Tait retention research
1.5–3×
what a subcontractor default typically costs against the original subcontract value
Surety industry data
A signed contract is a set of promises that only pay off if somebody administers them: obligations tracked, securities kept alive, retention accounted for, notices served on time, terms flowed down.
World Commerce & Contracting’s research puts the cost of weak contract management at 9.2% of annual revenue. In construction, where contract-administration failures sit among the top causes of disputes averaging tens of millions, signature is the halfway point rather than the finish line.
The leakage happens after award, and the reason is structural rather than lazy. The commercial expertise that negotiated the deal — procurement, legal — walks away at signature. Which is precisely when the obligations start biting.
World Commerce & Contracting puts the average loss at 9.2% of annual revenue, with top performers around 3% and laggards at 15–20%. On large capital projects the leakage reaches roughly 15%. And the biggest causes are entirely mundane: unauthorised and unrecorded changes, unmanaged clauses, missed renewals. Nobody loses 9% of revenue to a dramatic failure. They lose it to a diary nobody kept.
Construction adds its own sharp edges. UK government research found £3.2–5.9 billion held in retention at any one time, with around £240 million a year lost to upstream insolvency — and 44% of surveyed contractors had lost retention money that way. That is money already earned, sitting in somebody else’s account, disappearing when they fail.
And the instruments meant to protect everyone carry a brutal trap: an expired guarantee cannot be called. A project that runs six months late past its bond expiry — with nobody diarising the date — leaves you holding zero security at exactly the moment default risk peaks. The bond was not called. It simply stopped existing.
The securities ledger is a clock, not a filing cabinet. Performance bonds — the surety literature puts them typically in the 5–10% of contract value range, and in GCC practice they are almost always unconditional and on-demand, which makes them as good as cash to whoever holds them. Advance payment guarantees, which must reduce as the advance is recovered through certificates: leave one at full value and the subcontractor’s bank facility stays entirely blocked, tightening the very cash the guarantee existed to protect. And retention bonds. Every one of them carries an expiry date that will not match your actual progress. The disciplines are dull and they work: a diarised expiry ledger with 60- and 90-day alerts, extend-or-pay clauses, and release only once the defects period has actually closed.
The retention ledger is real money. Retention tracked per subcontractor, split into first and second moiety, against taking-over and defects-closure triggers. Orphaned retentions — held past their release trigger with nobody accountable — are a classic audit finding, an acid on subcontractor relations, and on the UK evidence a genuine insolvency exposure. Retention reform is a live regulatory question in the UK, with proposals to ban or ring-fence retentions under active consideration. Worth watching if you have UK exposure.
Notice compliance runs in both directions — and the asymmetry is the trick. Time-bar clauses extinguish claims that miss the window, and tribunals keep confirming they mean it. So the main contractor must serve its own notices up the chain and log the notices arriving from below. The detail that separates a good commercial team from an average one: subcontract notice periods should be shorter than the main contract’s, so that a claim arriving from a subcontractor can be passed upward before your own window shuts. Match the periods and you will eventually be time-barred upstream by a notice that arrived downstream exactly on time.
“Back-to-back” is an engineering task, not a phrase. A one-line clause saying the subcontract is “back-to-back with the main contract” fails routinely, because tribunals construe vague incorporation narrowly. Real flow-down means specific clauses actually incorporated: payment timing, variation valuation, liquidated-damages caps, suspension and termination triggers, notice periods. And variations must flow both ways — a main-contract variation without its mirrored subcontract variation means you absorb the delta yourself, quietly, and find out at final account.
Default has early warnings, and they are cheap. Labour deployment thinning on site. Sub-tier suppliers calling you — the main contractor — about bills your subcontractor has not paid. Requests for advance payment. Insurances lapsing. Quality and safety metrics sliding. A bank declining to renew a bond. Surety data prices the outcome: a subcontractor default typically costs 1.5 to 3 times the original subcontract value. Detection sixty days earlier changes the outcome, which is why these signals belong in a system rather than in the memory of whoever happened to take the call.
This is the one that catches international contractors, and it catches them expensively.
Conditional payment clauses — pay-when-paid, pay-if-paid — are effectively banned in the UK, with narrow exceptions, and restricted across much of the common-law world. But per law-firm commentary on UAE practice, they are valid and routinely upheld there, subject to limits around certainty and good faith.
So the same clause, in the same words, allocates payment risk completely differently depending on where you are standing. Anyone administering GCC subcontracts on UK instincts — or the reverse — is mispricing that risk, and will find out at the worst possible moment. Manage Gulf payment risk under Gulf law, not under imported assumptions.
The chains write themselves. The project is delayed. The bond expires, undiarised. The bank declines to renew it. You are now carrying zero security at peak risk, and nobody made a decision — a date simply passed.
Or: the advance payment guarantee is never reduced, so the subcontractor’s facility stays blocked, so their cash tightens — and you have caused the default the guarantee existed to prevent.
Or the worst one, the double bind: a missed notice time-bars your claim upstream, while the same event remains perfectly payable downstream. Margin eaten from both ends at once, by a deadline nobody was watching.
And retention never released — goodwill destroyed if you are lucky, money lost to somebody else’s insolvency if you are not.
Every subcontract and supply contract carries its obligations register, its securities with expiry alerts tracked against the actual programme rather than the original one, its retention ledger with release triggers, its notice log with the contractual clocks running, and its variation linkage back to the main contract.
The commercial position — commitments, securities, retentions, open notices, claims exposure — is one live picture per subcontractor and per project. Not twelve spreadsheets, each with a different truth, none of them wrong enough to be obviously wrong.
Securities are tracked against the actual programme, so a bond does not quietly expire while the project runs late.
Retention is accounted per subcontractor against its real triggers — released when due, and never orphaned.
Notice clocks run in both directions, with the downstream window deliberately shorter than the upstream one.
Distress is visible sixty days before it becomes a default that costs 1.5–3× the subcontract value.
Key dates, milestones, deliverables, caps and triggers extracted on award — the contract as a live record rather than a PDF in a folder.
Bonds, guarantees and insurances diarised with 60- and 90-day alerts, tracked against actual progress. An expired guarantee cannot be called.
The APG reduces as the advance is recovered — so a subcontractor’s facility is not left blocked by the instrument meant to protect them.
First and second moiety per subcontractor, against taking-over and defects-closure triggers. Orphaned retentions surface rather than settle.
Notices served and received, each with its contractual clock — and subcontract windows set shorter than the main contract’s by design.
Main-contract variations checked for their mirrored subcontract variation, because the delta you fail to pass down is the delta you absorb.
Key dates, milestones, deliverables, notice periods, caps and triggers — lifted out of the executed contract and into a register somebody actually reads.
Performance bond, advance payment guarantee, retention bond, insurances. Each with its expiry diarised and alerted at 60 and 90 days — against actual progress, not the original programme.
Specific clauses incorporated, not a one-line reference. Payment timing, variation valuation, LD caps, suspension triggers — and notice periods deliberately shorter downstream than upstream.
Notices served up the chain and notices received from below, with the contractual clock visible on each. A time bar does not care who forgot.
First and second moiety per subcontractor, against taking-over and defects-closure triggers. An orphaned retention is somebody’s money and your audit finding.
Labour, sub-tier supplier calls, advance-payment requests, lapsing insurances, bond-renewal refusals. Sixty days of warning is the difference between a managed handover and a 1.5–3× default.
Bonds, guarantees, insurances and licences tracked against actual programme dates rather than original ones — so the extension need is flagged while the bank still has time to act, instead of discovered after the instrument has lapsed.
Events detected across the record — instructions, delays, disruptions — checked against notice obligations in both directions, with the deadlines visible and draft notices prepared for commercial review.
Main-contract variations with no mirrored subcontract variation are flagged, and subcontract terms are checked against the back-to-back positions you actually intended. The delta you do not pass down is the delta you keep.
Payment-application patterns, site records and compliance signals correlated per subcontractor, so the distress pattern surfaces while intervention is still cheap. A default costs 1.5–3× the subcontract value; the signals cost nothing.
The engineer’s judgment stays in charge; the AI removes the latency and the blind spots.
One live commercial position per subcontractor and per project: commitments, securities with their expiry clocks, the retention ledger against its release triggers, open notices in both directions, and claims exposure. Distress indicators trend per subcontractor. Exportable for commercial governance — and structured so the securities register is a diary rather than a drawer.
The gap between the value you negotiated and the value you actually realise. World Commerce & Contracting research puts it at 9.2% of revenue on average, driven mostly by unrecorded changes, unmanaged obligations and missed dates after signature — not by anything dramatic.
A performance bond secures performance — the surety literature puts them typically in the 5–10% of contract value range, and in most GCC practice they are unconditional and on-demand. An advance payment guarantee secures the advance and should reduce as it is recovered through certificates. A retention bond replaces cash retention, freeing the subcontractor’s cash flow. All three share one failure mode: expiry before release.
It becomes uncallable. Worthless. Delayed projects routinely outlive their securities unless somebody extends them — so the mitigations are diarised expiry tracking against actual progress, extend-or-pay clauses, and renewal demands served well before the date.
It depends entirely where you are standing. They are effectively banned in the UK and restricted in many jurisdictions — but per law-firm commentary on UAE practice they are valid and routinely upheld there, subject to limits around certainty and good faith. Manage Gulf payment risk under Gulf law, not under imported assumptions.
Read the full answerSpecific incorporation. Payment terms, variation valuation, liquidated-damages caps, suspension and termination triggers, and notice periods — set shorter downstream than upstream, so a claim from below can be passed above before your own window shuts. A vague one-liner fails when it is tested, because tribunals construe loose incorporation narrowly.
Read the full answerThinning labour on site. Sub-tier suppliers contacting you about unpaid bills. Requests for advance payment. Lapsing insurances. Declining quality and safety metrics. A bank refusing to renew a bond. A default typically costs 1.5 to 3 times the subcontract value — the signals are very much cheaper than the outcome.
Read the full answerRelated answers
In the guides
Terms defined here
Zepth is the construction project delivery platform — it runs construction, procurement and asset management on one record, and does the work: reading the drawings, reviewing the submittals, matching the invoices and flagging the risks, with a human sign-off on anything consequential.
A short, tailored walkthrough on your real workflow — no generic demo.