Build a Cost Value Reconciliation Spreadsheet for Construction

A cost value reconciliation spreadsheet is a monthly, package-level workbook that compares gross value certified to the client against cost incurred and accrued on each subcontract package and direct-cost category. The commercial QS runs it to surface margin slippage at the package row before it compounds into a project-level loss.

The structure follows from that purpose. Rows are subcontract packages (groundworks, frame, MEP, finishes, and so on) plus direct-cost categories: materials, plant, labour direct, preliminaries. Columns track budget, period and cumulative value, period and cumulative cost, accruals, adjusted cost, margin, and forecast final cost. The cells get populated each month from a fixed cast of source documents: the subcontractors' applications for payment, the client interim certificate, the purchase ledger filtered by job and cost code, plant hire dockets, labour timesheets, and an accruals workpaper that closes the gap between cost incurred on site and cost posted to the ledger.

What follows is the source-document-to-row walk-through — which document each cell of the matrix is sourced from, how it is pulled, where the trap is, and what the resulting margin signal means operationally.

---

Inside the package-level matrix: rows, columns, and the budget column

The matrix has two dimensions: a row for every cost centre that earns or spends money on the project, and a column for every figure the QS needs in order to read margin position by row.

Rows. Every subcontract package gets its own row at whatever granularity procurement set up — groundworks, substructure, frame, envelope, MEP, fit-out, externals, whatever the procurement schedule named. Below the subcontract block sit the direct-cost categories: materials (on-cost from supplier invoices that aren't covered by a subbie's package), plant (hired plant, internal plant), labour direct (in-house operatives, agency labour where it isn't carried within a subcontract), and preliminaries (the project's own overheads: site management, welfare, supervision, scaffold, hoardings, security, temporary services).

The package row is the unit of analysis. A project-level margin can read healthy while one package quietly bleeds and another over-recovers, and a CVR that only reports at the project level cannot see that. Margin slippage hides at the package row, never at the project row, and that observation is the recurring frame for everything that follows.

Columns. Each column has a precise meaning, and the workbook is only as trustworthy as the discipline behind those meanings.

• Budget. The package's priced value at procurement. For subcontract packages, this is the agreed subcontract sum; for direct-cost categories, it is the priced bill of quantities allowance for that category. This is your fixed reference point; it does not move month to month unless a contract amendment makes it move. The upstream step is taking the priced BOQ into the workbook in the first place — a separate exercise that needs its own discipline. The pre-CVR step of convert a tender BOQ PDF into a priced Excel sheet is where this column is sourced.
• Period value. The increment of gross certified value this month: cumulative gross certified value at this cut-off minus cumulative gross certified value at last cut-off.
• Cumulative value. Gross certified value to date for this package row, taken from the latest client interim certificate (or its package-level breakdown).
• Period cost. Cost incurred this month against the package row, before accrual adjustments.
• Cumulative cost. Cost incurred to date — the cumulative ledger total filtered by job number and cost code, plus any prior-period accruals carried forward.
• Accruals. Cost incurred but not yet posted to the ledger, plus any subbie liability not yet on a paid invoice. This is where dockets, timesheets, and goods-received-not-invoiced positions land.
• Adjusted cost. Cumulative cost plus accruals plus prelims and over-measurement adjustments — the figure margin is read against, not the raw cumulative cost.
• Margin. Cumulative value minus adjusted cost, read in absolute terms and as a percentage by package and at project level.
• Forecast final cost. Adjusted cost projected to package completion, typically built from cumulative cost plus a remaining-cost forecast tied to the programme.

Budget is locked once procurement closes the package, with a clear note where it has changed and why. Every other column moves month to month with the cut-off.

Most main contractors still run this workbook in Excel. Vendor CVR modules exist and have their place, but a structured Excel template that the QS can hand-edit, the commercial director can read at a glance, and any project accountant can tie back to the ledger remains the practical deliverable for the majority of teams. That is why this article works through the workbook structure rather than a tool's interface.

Value side: client interim certificate to gross certified value

The value column has one authoritative source: the client interim certificate under JCT, or the AfP to client (with the project manager's assessment) under NEC. That document gives you the cumulative gross certified value at the cut-off date. Subtract last month's cumulative figure and the period value column populates itself.

The drill into per-package value rows depends on the certificate's format. Where the certificate carries a package-level breakdown — many JCT interim certificates do, particularly on bigger projects where the contract sum is itemised by element or trade — copy the package figures across directly. Where the certificate aggregates packages or runs at the level of broad work groups, the QS apportions using the internal valuation: take the certified total, distribute it across packages in the proportions the internal valuation supports, and document the apportionment so it can be defended at sign-off.

The internal valuation itself sits alongside the certificate in any well-run pack. The QS's internal value of work done at the cut-off date typically leads the client certificate by a month — the certificate reflects work measured up to the previous valuation period, while the internal valuation captures site progress to the current cut-off. The internal valuation also tends to include items the client has not yet certified: re-measurable work the QS believes will hold, late instructions waiting on a variation order, claims that have not yet flowed through the certified figure.

The workbook should carry both numbers, and the QS should be deliberate about which one feeds margin reporting. Most teams report against the certified figure as the conservative position and use the internal valuation as a forward indicator — a leading signal that next month's certificate will catch up or, where it won't, that a value-recovery exercise is needed before the gap becomes a write-off. A few teams report against internal valuation; either approach works, but the choice has to be consistent month to month or the trend lines lose meaning.

This is the half of the reconciliation that a sloppy CVR gets wrong first. Value vs cost reconciliation on a construction project only has analytical force when the value side is defensible, so the same care that goes into the cost-side ledger pull belongs on the value side too. If the value side isn't defensible, the margin number isn't defensible either.

The trap to flag — published material on CVR consistently muddles it — is the conflation of "value of work done" with "gross certified value". They are not the same number. Internal value of work done can sit above gross certified value (work physically complete but not certified yet) or below it (an over-claim from an earlier period that the certificate has corrected downward). The matrix carries one or the other; pick a definition for the value column and stay with it. Mixing the two month to month is what produces a margin chart that wanders.

Over-measurement, contractual-claim caveats, and the question of when to provision against a certified figure that the QS does not believe will hold are real concerns on the value side, but they sit a layer above this source-to-row mapping and have their own treatment further on.

Cost side: pulling incurred cost from the purchase ledger and assessed AfPs

The cost-side spine is the purchase ledger filtered by job number and cost code. Run the ledger report at the CVR cut-off, filter to the project's job number, and group the postings by cost code. The cost-code structure should map one-to-one to the matrix: each subcontract package has its own cost code or code range; materials, plant, labour direct, and preliminaries each have theirs. Cumulative cost in every row of the workbook comes from this filtered ledger before any accrual adjustments are applied.

That mapping only works if cost coding upstream is disciplined. Every supplier invoice, every subbie payment, every plant hire invoice, every internal labour journal has to carry the right job number and the right cost code, posted at AP not corrected after the fact. When cost coding is sloppy, cost lands in the wrong package, one row over-reports while another under-reports, and the package-level margin signal is masked. The upstream discipline of getting AP to consistently code construction supplier invoices to job and cost code is what makes the CVR's cost side trustworthy in the first place; if that discipline isn't there, the QS spends the cut-off week chasing miscodings rather than reading margin.

The subbie liability line is built differently. The cost figure for each subcontract package is the cumulative assessed value of all the subbie's AfPs to date — not the value paid, and not the value applied for. The two distinctions matter.

• Assessed, not applied. Use the value the QS assessed on each AfP, after pay-less adjustments where they applied. Some teams default to the gross applied figure because it's easier to lift off the front sheet of the application, but that overstates the package's cost-side position with figures the contractor never agreed to. The right number is what the team genuinely owes the subbie, which is the assessment, and the assessed figure is the one that survives a director-level conversation. If the upstream AfP discipline is shaky, the article on how to assess subcontractor applications for payment covers what the assessment process should look like.
• Assessed, not paid. Use cumulative assessed value, not cumulative paid. Payment lags assessment by net 30 to net 45 days under most subcontract terms. Using paid would systematically understate the cost side by a month or more and produce a margin number that is flattering today and embarrassing next month when the payment runs catch up.

That gives a clean cost row per package. The reconciliation between assessed AfP value and the ledger position is the next step. The ledger shows what's been posted: subbie payments processed and any subbie cost the project accountant has formally accrued. The assessed figure shows what is owed. The CVR cost row uses the assessed figure; the gap between assessed and posted feeds the accruals workpaper that the next section walks through.

Spot-checking against ledger miscoding is the one check the QS should not skip before treating the matrix as final. The most common cost-side trap is a supplier invoice coded into the wrong package — a frame subcontractor's labour-only invoice posted against the externals package because the AP clerk didn't know the difference, a hire invoice for the tower crane coded to plant when it was meant to sit in prelims. Either inflates one row's cost and starves another's, and the package margin reading goes wrong in two places at once. Tie the ledger's package totals back to the source documents in spot-check fashion: delivery notes against the materials cost row, hire dockets against plant cost, AfPs against the subbie row. Ten minutes of tying out the largest postings catches almost all the miscoding before sign-off.

The operational reality on a typical UK main-contractor project is that this section's work involves pulling figures off heterogeneous PDFs at scale every month: dozens of subbie AfPs, hundreds of supplier and merchant invoices coded against the project, plant hire dockets, the client interim certificate referenced earlier. Doing that by hand — opening each document, locating the figure, retyping it into the workbook or the assessment template — is the QS team's biggest single time sink in any given month, and it is where avoidable transcription errors creep into the matrix. When extracting figures off these documents has become the bottleneck on the cut-off week, the team's leverage point is to automate construction document extraction so the cut-off week can be spent reading the matrix rather than retyping its inputs.

Filling the lag gap: dockets, timesheets, and the accruals workpaper

The ledger only knows what AP has posted. At the CVR cut-off, plant on hire that hasn't been invoiced yet, materials delivered against an open purchase order, in-house labour worked but not yet journalised — none of it shows up in the ledger filter the cost side relies on. Without an accruals workpaper that closes that gap, the matrix under-reports cost across multiple packages and the margin column prints a falsely flattering number that month-end accruals will quietly correct out of view.

The source documents that feed the workpaper are the documents the project team already produces on site, not new ones manufactured for the CVR.

• Plant hire dockets at the gate, or the on-hire register the plant manager keeps. Every piece of hired plant on site at the cut-off date that hasn't been invoiced needs an accrual at the going rate for the days on hire since the last invoice was raised. Where the on-hire register is up to date, the workpaper is straightforward; where it isn't, the QS needs an hour with the plant manager to bottom it out before the cut-off.
• Labour timesheets for direct labour worked but not yet journalised. The pinch point is in-house operatives where payroll is run on a different cycle from the CVR cut-off; agency labour is usually invoiced weekly and posts close to real time, but in-house wages may lag by a fortnight. Accrue the unposted hours at the standard internal cost rate.
• Materials delivery notes against open POs. Goods-received-not-invoiced is the classic cost accrual: the delivery note proves the cost was incurred, the missing supplier invoice means the ledger doesn't see it yet. Reconcile open POs against goods received, accrue the difference at the PO rate.
• The AfP-versus-ledger gap from the previous section. Cumulative assessed AfP value minus what the ledger has posted for that subbie equals the subbie-side accrual. This is normally the largest single line on the workpaper.

The accruals workpaper itself is a simple structure: one row per accrual reason, with columns for the source document reference (docket number, timesheet week, delivery note number, AfP number), the package allocation, the value, and a brief justification. Total accruals roll up to the accruals column in the matrix, and the workpaper is the audit trail behind that column. At sign-off, the commercial director won't read every row, but they will sample — pull a docket, pull a delivery note, see whether the workpaper reconciles. The workpaper has to be defensible at that level of scrutiny.

Materials accruals are the most common slippage source on a typical CVR. Builders' merchant invoices for timber, fixings, ironmongery, plasterboard, and similar consumables run on monthly statements that arrive mid-month after the cut-off, which means the bulk of a month's materials cost is invariably sitting in goods-received-not-invoiced at the moment the CVR is being assembled. Building the workpaper line for materials accruals from the delivery-note pile means having a fast way to read figures off those merchant documents — a skill set that ties directly to the broader exercise of how to extract line items from UK builders' merchant invoices once those invoices arrive and are reconciled against the accruals raised.

The reverse trap is double-counting. If the QS accrues a plant hire docket at this cut-off and the supplier invoice posts to the ledger before the next CVR cycle, the accrual must be reversed at the start of next month, otherwise the cost is in the matrix twice — once in the cumulative ledger, once carried forward in accruals. Standard practice on a well-run pack is to reverse all accruals on day one of the new cycle and rebuild from current dockets, current open POs, and the new month's AfP-versus-ledger gap. The reversal-and-rebuild discipline is what keeps the accruals column honest period to period.

External preliminaries adjustment

Preliminaries are project overheads — site management, welfare, supervision, scaffold, hoardings, security, temporary services, plant kept on site for the project team's own use rather than for a specific package, and the cost of running the site office. They don't sit inside any single subcontract package and they don't fit neatly into a direct-cost line either. The prelims row in the matrix carries them as their own category, and most of the time the row populates itself from the ledger like any other cost-coded line.

External preliminaries adjustment is the slice of prelims that the ledger pull does not capture cleanly. Two scenarios produce it, and both turn up frequently enough that the QS who skips them is consistently misreading the prelims line.

The first is recharged prelims. A site shares a tower crane with the project next door, the welfare cabin is invoiced centrally and split across four jobs, the regional commercial team's quantity surveyor charges time across multiple projects, the firm's insurance is allocated to projects on a percentage basis. Whenever a prelims cost is incurred against this project but billed against another job number — or, equally commonly, billed against this project but properly belonging to another — the recharge has to be adjusted in or out before the prelims row reads true. The mechanics are straightforward when the project accountant runs a clean recharge schedule each month: take the schedule, post the recharges as adjusted-cost entries against the prelims row (positive for cost coming in, negative for cost going out to another project), document the basis. Where the recharge schedule is informal or runs on an ad-hoc spreadsheet, the QS needs to chase it down before the cut-off rather than after — a recharge picked up two months late distorts both projects' margin trends in ways that are slow to unwind.

The second is the prelims accrual that doesn't match value progress. Prelims tend to be roughly fixed in the rate at which they burn — site management cost, welfare, supervision, the site office — running at a relatively constant cost per week across the programme. The value side certified by the client typically tracks measured work, not time, so the rate at which value is certified rises and falls with the work-package mix on site. Mid-programme that's fine, because cumulative value broadly tracks cumulative cost on a good project. Toward project completion, though, certified value tends to plateau — the bulk of the work is in, the snag list and final-fix activities are slow to convert into a measure — while prelims keep burning at the same rate as the team demobilises. The prelims row's adjusted cost trends upward against a value side that has flattened, and the package matrix needs a forward-looking prelims accrual to surface the closing-stage drag rather than waiting for it to land at final account.

The mechanics in either scenario are the same. The adjustment lands in the adjusted-cost column for the prelims row, either folded into the row's adjusted-cost figure or shown as a separate adjustment row underneath if the team values the visibility. Either way, the calculation and source need documenting on a workpaper alongside the matrix: the project accountant's allocation method for cross-project recharges, the programme-versus-cost-curve analysis for time-based accrual.

The pitfall the published material consistently understates is the reverse — under-stating external prelims so that the prelims row reads more flatteringly than reality. A workbook that posts prelims at ledger value alone, with no recharge adjustment and no closing-stage accrual, prints a margin that the next quarterly review will catch and, failing that, that final account exposes. Surface the adjustment now, document the basis, and the prelims row stays defensible whether the closing position is good news or bad.

Over-measurement and contractual claim caveats on the value side

The value-side section earlier took the gross certified value off the client interim certificate and treated it as the value column figure. That figure is correct as a record of what has been certified, but the QS reading the matrix at margin-review level needs to know whether to read the certified figure as is or apply caveats that adjust the package margin before reporting. Two caveats matter in practice — over-measurement and contractual-claim adjustments — and the published material on CVR mishandles both.

Over-measurement. A quantity certified to the client is over-measured when the QS believes the certified quantity will not hold up at final remeasure. It happens for unspectacular reasons. A certificate gets settled on assumed quantities pending site verification and the verification later reduces them. Earlier valuations stack optimistic measures on a fast-moving package — concrete poured, formwork struck, façade panels installed — and a careful re-measure run by the QS later in the programme identifies that the cumulative certified quantity ran ahead of what is actually in the works. A subbie's package was certified on a percentage-complete basis and the percentage was generous against site reality.

The certified figure stays in the value column, because the certificate is the contractual record. The over-measurement caveat sits alongside it as an over-measurement provision: a negative entry in an adjustment line within the value side, or a footnoted reduction read against the package row. The package margin is read after the provision is applied, not before. The basis for the provision needs documenting — a re-measure exercise, a quantity-surveyor sample audit of the package, a project manager's flag — and so does the recovery scenario, in plain terms: the QS expects the provision to clear in full at the next valuation, expects it to reduce by half, expects it to crystallise as a value reduction at final account. Trend the provision month to month and the matrix reads more honestly than the certificate alone.

Contractual claim adjustments. Claims work in both directions. Loss-and-expense claims, prolongation, disruption — claims the contractor has against the client that the certificate hasn't fully captured but that the QS knows are in train — sit on the value side as positive provisions, with the basis documented and a recovery scenario noted (likelihood of agreement, expected timeline, likely value if settled at less than the claimed sum). Conversely, claims being made by the client against the contractor — defects, delay damages, contra-charges for shared facilities — show up as negative provisions on the value side because they will reduce certified value when settled. The mechanics are identical to over-measurement: provision in the matrix, basis on a workpaper, recovery scenario noted, trend month over month.

Variations sit alongside both. Variation values that have been agreed and certified flow through the value column as normal — they are part of the certified figure already. Variations that have been instructed but not yet agreed are the awkward case: the work is being done at the QS's day-rate or assessed value, the cost is hitting the cost side, and the value side has nothing to show for it because the variation order hasn't been issued and the certificate hasn't picked it up. Carry these as a value-side provision pending agreement, exactly mirroring the over-measurement logic in the opposite direction. They become certified value when the variation order lands; the provision unwinds at that point.

The pitfall is skipping the caveats and reading the certificate as gospel. A workbook that prints a margin number on uncaveated certified value tells the commercial director that the project is making more money than the QS believes it is — for over-measured packages — or less than it actually is, for packages with unagreed variations. Either way the matrix loses defensibility at the moment it most needs it, which is the month-end commercial review when the director asks why the margin has moved. The provisions are how the QS produces a margin number that matches their professional judgment of the position rather than the contractual record alone.

---

Reading margin slippage at the package row

A populated CVR matrix is only useful when somebody reads it. The reading is not at the project level — the project margin is an aggregate that hides as much as it shows — but at the package row, scanning each row's margin column against its budget and watching the trajectory month over month. A CVR pack that doesn't surface package-level diagnosis is a pack the commercial team can't act on.

Four slippage patterns turn up regularly enough to be worth recognising on sight. Each has its own source-document evidence and its own operational response.

Cost overrun in a single package. Adjusted cost trending above budget on one package, with the overrun building period over period rather than appearing in one anomalous month. The source-document evidence is in the assessed-AfP series for that package — the subbie's monthly applications climbing faster than the original subcontract sum agreed at procurement — or in the package's ledger detail, where extra orders or late-running variations are being absorbed as cost without a corresponding value claim. The operational response is some combination of three things: re-measure the package against what is actually built to test whether the assessed values are warranted; issue a pay-less notice on the next AfP if the subbie's application is unsupported; raise variation-order back-up if the cost reflects genuinely instructed change that hasn't been picked up on the value side yet. The last of those moves the cost into the value column too, which often resolves the slippage without further action.

Value under-claim. Period value for a package is materially smaller than period cost — work is being done and paid for, and the value side isn't catching up. The source-document evidence is the on-site progress for the package versus the percentage that the certificate has credited. Either the QS missed the work in the internal valuation that fed this month's AfP to the client, or the certificate didn't give full credit to what the internal valuation included. The operational response is to rebuild the internal valuation against site reality and push the missing measure into next month's AfP to client. Closing the gap takes a month, but the matrix carrying a known under-claim with the recovery action documented reads better than a matrix that prints a depressed margin without explanation.

Scope change unrecorded. A package's cost is rising without any corresponding value adjustment, and there's no variation order on file. The source-document evidence is in the project record — the foreman's diary, the project manager's instructions log, the subbie's covering letter to the AfP describing work outside the original scope, RFIs against the package that have been answered with effective instructions. The operational response is to regularise the variation: get the variation order issued, value the work, get it on the next AfP to client, and update both sides of the matrix. The longer it stays unregularised the harder the value recovery becomes, because the contractual paper trail goes cold.

Preliminaries drift. The prelims row's adjusted cost is climbing faster than value progress, particularly on the back end of the programme. The source-document evidence is the prelims-resource schedule against the construction programme: site management, supervision, scaffold, hoardings, security, temporary services running at their normal rate while the package activity that is generating the value side has slowed. The operational response is to re-baseline the prelims forecast and surface the drag to the commercial director early, while there is still time to adjust the demobilisation programme or to take prelims-cost-saving action.

The matrix's role in escalation matters at least as much as its role in diagnosis. The QS's job is to surface slippage with the source-document evidence in hand: the assessed AfP series, the on-site progress photographs, the foreman's diary entries, the prelims resource schedule. A matrix that's tied back to its source documents is a matrix that survives a director-level conversation; a matrix that prints package margin numbers without the workpapers behind them collapses the moment the director asks why. The package CVR a commercial team can take into a margin review is the one with the source documents stacked alongside it.

The discipline matters at the firm level, not just the project level. Construction was the UK industry with the highest number of insolvencies in the 12 months to November 2025, with 3,950 insolvencies, 17% of all cases where the industry was captured, according to the UK Insolvency Service's December 2025 company insolvency statistics. The contractors that survive aren't the ones that avoid loss-making projects entirely — every business in the sector takes on projects that go wrong — but the ones that catch package-level slippage in month one rather than month nine, while there's still operational room to act on it. That is what the CVR is for, and that is the standard the matrix has to meet every cycle.

Audit trail, sign-off, and the tie-out to month-end accruals

A CVR is only as good as its tie-outs. Every figure in the matrix should trace to a source document or a referenced workpaper: the certificate for value, the assessed AfPs for subbie liabilities, the ledger by job and cost code for posted cost, the accruals workpaper for the cost-versus-ledger gap, the recharge schedule and prelims allocation for external preliminaries, the over-measurement provisions for the value-side caveats. A matrix without those tie-outs cannot be defended at sign-off, and a CVR that cannot be defended drives no decision worth taking.

The sign-off chain typical of a UK main contractor is short and well-understood. The project QS owns the matrix, the workpapers, and the source documents behind both. The commercial manager reviews and challenges — sense-checking package trajectories, sampling the workpapers, pushing back where the QS's reading of slippage doesn't square with what they hear from the project team. The commercial director signs off. The director's review focuses on package-row margin trajectories, the slippage patterns the previous section walked through, and the forecast final cost. They are not re-checking individual ledger postings; they are testing whether the QS's margin position is defensible against what they know about the project's operational reality.

Downstream of the matrix, the project accountant takes the CVR's accruals total and the cumulative-cost-versus-ledger-posted figure into the management accounts. The accruals position is journalised as work-in-progress (or as a contract asset, depending on whether the firm runs FRS 102 or IFRS 15 accounting policy), and the project's contribution to the period's reported profit follows from the CVR margin position. That translation is the project accountant's, not the QS's; the CVR figure is the input. Teams whose AP and commercial functions automate invoice processing across a construction business typically find the handoff faster too, because the cost-side data feeding both the ledger and the CVR is already structured rather than re-keyed off paper.

Further out, at project close, the final CVR position becomes the starting point for final-account preparation against the client and for closing each subcontract package's final account. Every monthly CVR's discipline — the source-document tie-outs, the accruals reversed and rebuilt cleanly each cycle, the over-measurement provisions trended and unwound deliberately, the prelims recharges agreed in real time rather than backdated — is what makes the final accounts defensible when they land. A project's twelve monthly CVRs are, in aggregate, the audit trail behind its final account.