One project, one commercial truth

Silos destroy causation before they destroy margin

The typical project organisation contains all the expertise needed to understand a change and none of the routines needed to combine it. Operations see that a workface is unavailable. Planning moves crews and edits the next update. Commercial issues a notice and opens a variation. Cost control reports an adverse variance. Everybody acts reasonably inside a professional boundary, and weeks later the outputs do not reconcile.

The planner recorded a new sequence without the operational reason for it. The quantity surveyor valued standing time from a subcontractor quotation without knowing the crew was reassigned. The site manager described late access without identifying which scheduled activity was due to start. Finance accrued a recovery without segregating the cost. Senior management now has four versions of one event: operational, contractual, temporal and financial.

That fragmentation damages more than a future claim. It delays mitigation, lets notices drift away from facts, hides contractor performance problems and produces unreliable forecasts. The commercial report shows a large entitlement while the programme asserts no compensable critical delay and the ledger holds a fraction of the attributable cost. Apparent margin has replaced commercial truth.

An integrated team does not require every discipline to agree. It requires each proposition to be visible and testable. Operations may say the event changed the work; planning may say it was not then critical; commercial may say the direct variation cost is recoverable but the prolongation is not; cost control may say only part of the labour claim reconciles. That disciplined disagreement is worth more than artificial consensus.

Four professions, distinct accountabilities

Integration fails when it becomes role dilution. The disciplines share an event model and keep clear accountability.

• Operations own physical fact and mitigation: what was planned and performed, where resources went, what constraint arose, what instruction was received, how the team responded. Operations do not decide legal responsibility in a diary.
• Planning owns time integrity: the baseline and update history, verified actual dates, explained logic, calendar and constraint changes, current and near-critical paths, and the line between forecast and retrospective opinion. A planner does not mechanically insert employer events into a programme handed over by commercial.
• Commercial management owns the bargain and the valuation process: mapping facts to clauses, protecting notice requirements, obtaining instructions, valuing changed scope, segregating event cost, updating the forecast, developing submissions. It must resist claiming every cost variance as compensation.
• Cost control and finance own financial lineage: cost-code discipline, ledger reconciliation, accrual rules, commitments, forecast to complete. Their systems prove cost exists. Commercial and operational evidence explains why.

Document control protects the controlled record. Legal counsel advises on governing law, privilege, disclosure and procedure. Quality, design, procurement and risk contribute sources. The project director owns the integrated account and resolves priorities.

Experts sit deliberately outside all of it. A delay or quantum specialist may audit the system, identify missing data and explain analytical requirements. Once appointed to give independent evidence, their overriding duty is to the court or tribunal, not to defending the project’s commercial forecast. Project advocacy, factual witness evidence and independent opinion have to stay separable.

The cadence

Daily: the production truth

At shift close the supervisor records resources, quantities, locations, constraints, instructions, downtime, rework, mitigation and source photographs against the relevant activities and event identifiers. The planner or field engineer validates actual starts, finishes and progress. Potential events are flagged within twenty-four hours.

The record must distinguish attendance from deployment, and deployment from productive output. Twelve electricians on site is attendance. Eight on containment, two awaiting access and two correcting defective work is allocation. Metres installed per labour hour is output. Only the full sequence supports productivity analysis, which is why disruption claims built on attendance sheets fail.

Weekly: the event-control meeting

Commercial, operations, planning, cost control and document control review the material events and decide, for each one: what is known, disputed and missing; which work and programme activities are affected; whether notice or further particulars are due; what mitigation is available, authorised and costed; how resources, quantities and costs will be segregated; whether criticality or milestone exposure has changed; who owns the next decision and by when; and whether the event is closed, monitored, submitted or escalated.

Minutes record the reasoning, including minority views and uncertainty. The objective is not to declare entitlement in thirty minutes. It is to stop the factual and contractual positions diverging.

Monthly: the integrated project account

The monthly cycle freezes a controlled data date, and programme, cost report, change register, risk register, progress measurement and executive forecast all reconcile to it. The planner explains movement since the last update. Commercial reconciles submitted, assessed, forecast and received amounts. Cost control confirms incurred and committed cost. Operations validate the remaining sequence and the productivity assumptions.

Management receives scenarios rather than one aspirational number: base forecast, realistic recovery and downside exposure; employer-risk, contractor-risk and unresolved time; gross cost, submitted claim and independently assessed recoverability, with evidence maturity and notice compliance alongside.

Quarterly: independent challenge

A senior reviewer or external specialist takes a sample of events and follows them from source record to board forecast. This is neither audit theatre nor an early attempt to draft expert reports. It asks three things: can the project’s assertions be reproduced, are the programme and cost logic coherent, and have the uncomfortable facts been retained.

A worked example: fragmented access to a services corridor

This is a composite illustration of a recurring scenario, not a reported case, and no legal outcome is implied.

The approved method assumes a mechanical subcontractor receives 240 metres of clear corridor, letting one continuous crew install supports, duct and pipework in sequence. The employer releases the corridor in small, changing sections because preceding civil and design work is incomplete. Crews mobilise, stop, relocate and return. Completion may or may not move. Productivity certainly deteriorates.

The siloed response produces a weak claim. The diary says access restricted. The programme shows the overall MEP activity continuing. The subcontractor submits a lump-sum disruption figure. Commercial forwards it as a variation. With no quantity, resource or location history, the respondent answers that the contractor chose an inefficient sequence and was responsible for coordination, and there is nothing in the record to contradict that.

The integrated response starts from the planned condition: tender method, baseline logic, access schedule, planned crew and production rate establish the expected continuous workface. Daily records then capture each released segment by chainage or grid, the time of release, crew allocation, quantities installed, moves, idle hours and rework. Photographs are located and timestamped. Instruction and request histories explain why sections were unavailable. The planner shows how fragmented releases affected logic and whether float or resequencing absorbed the delay. Cost control segregates additional supervision, plant moves, overtime and subcontractor cost. The notices evolve with the evidence instead of asserting an unparticularised delay.

That analysis can produce three separate answers: direct change cost for additional moves or altered scope; disruption cost for proven productivity loss, potentially using comparable unaffected corridor sections; and prolongation only if a compensable event is shown to extend the relevant project period under the contract and the law.

It may also expose contractor risk: poor crew balancing, defective work, late procurement. Stripping those hours out lowers the headline number and raises the reliability of what remains. Commercial maturity is not maximising the first number. It is knowing which number survives examination.

The common data model

A team cannot integrate through meetings while its systems describe the project differently. The programme work breakdown, cost breakdown, location breakdown and organisation breakdown are mapped at mobilisation. They need not be identical, because construction software and finance systems have different jobs. They must be translatable, and a controlled dictionary should define activity, location, work package, cost code, event, notice, instruction, risk and document identifiers.

The event identifier is the join. It connects the observation to the notice, the programme narrative, the schedule fragment, the resource allocation, the quantity, the cost transaction, the risk and the decision. Where a transaction relates to several events the allocation rule is stated. Where allocation is not yet possible the cost is held as unresolved rather than pushed into a favourable code, which is the small discipline that keeps a cost model honest.

Commentary Published project-controls material from the large consultancies supports this integrated view of schedule, cost, risk, change and performance, and platform vendors make a parallel case from the records side. Both are commentary with a commercial context, including ours. Governance, configuration and behaviour decide whether integration is genuine; procurement decides very little of it.

Controls that expose uncertainty rather than conceal it

Cost-value reconciliation should separate four layers: contract value, forecast final account, actual and forecast cost, and risk-adjusted recoverability. Compensation events and claims should not be booked at their asserted value merely because a notice exists. Each material event needs an entitlement view, a causal view, a time view, a quantum view and a collection view.

A simple evidence-maturity scale helps: observed, where the source fact is recorded and responsibility is unassessed; protected, where notice and the relevant records are in place; analysed, where time and cost consequences have been tested against alternative causes; submitted, where coherent particulars and valuation have been issued; and determined, where the event has been agreed, assessed or decided subject to identified challenge.

That scale should never be converted mechanically into revenue recognition or legal probability. Its purpose is to separate a live site issue from an auditable recovery case.

Measure leakage too. If an instruction is recorded but cost codes open six weeks later, attributable cost has already mixed with base work. If a subcontractor claim arrives without shift allocations, the main contractor has inherited an evidence problem. Downstream contracts, site instructions and payment processes should require the same event structure, without imposing administration out of proportion to the risk.

Planning controls that preserve the analysis

Schedule quality is behavioural as much as technical. Preserve native baseline and update files, approved and rejected versions, calendars, coding dictionaries, progress rules, narratives and change logs. Actual dates reflect evidence rather than being adjusted to improve the model. Logic changes after the data date get an explanation. Out-of-sequence work and constraints get transparent treatment.

Report near-critical paths, not only the current longest path: a two-day movement that is immaterial today becomes critical after an access or procurement change. Treat float according to the contract and the governing law rather than presuming it belongs to one party. Concurrency needs precise facts and a jurisdiction-specific analysis, and the SCL Protocol, influential as it is, does not replace the applicable legal test.

Operational validation is indispensable. A mathematically valid path can be physically implausible. Planners should walk the work, compare quantities and understand methods; operations should not demand that the programme reproduce a preferred narrative. The monthly update is a controlled model of current knowledge, not a claim exhibit.

Nine measures for executive assurance

A board needs indicators that test capability rather than reward claim inflation. These work once definitions and thresholds are set project by project.

1. Percentage of material events recorded within twenty-four hours.
2. Notice and continuing-particular compliance against contractual deadlines.
3. Completeness of labour, plant, quantity, location and activity allocation.
4. Percentage of programme movement supported by an approved narrative.
5. Value of event-coded cost reconciled to the ledger.
6. Age and value of unresolved events.
7. Share of claimed value at each stage of the evidence-maturity scale.
8. Proportion of contractor-risk and neutral events recorded to the same standard as employer-risk events.
9. Results of the quarterly independent challenge: how much of the sample could be reproduced from source.

None of these measures the size of the claim, which is the point. They measure whether the project can explain itself.

The executive conclusion

One project, one commercial truth is not a slogan about alignment. It is a governance requirement. The same event has to mean the same thing in the diary, the programme, the notice and the ledger, and the disciplines that own each of those have to meet often enough that divergence is caught in days rather than years.

Do that and the claim, if there is one, is a controlled aggregation of decisions the project already understood. Skip it and the claim becomes an act of reconstruction, performed by people who were not there, against an opponent whose only job is to find the joins.