Section 20 Major Works Invoice Reconciliation Guide

For qualifying works, Section 20 consultation matters when any one leaseholder's contribution will exceed GBP250, and for qualifying long-term agreements the trigger is more than GBP100 per tenant per accounting period. In practice, section 20 major works invoice reconciliation means matching each contractor invoice to the consulted scope, the incurred date, the demand timeline, and the lease apportionment schedule so recoverability and year-end service-charge reporting can be defended. That threshold is reflected in the Service Charges (Consultation Requirements) (England) Regulations 2003, which set the Section 20 trigger at more than GBP250 per tenant for qualifying works and more than GBP100 per tenant per accounting period for qualifying long-term agreements.

That matters because the finance problem starts after consultation, not before it. Once the Statement of Estimates has gone out and works begin, the live question is whether each invoice still belongs to the works that were consulted on, whether it can be charged through the service charge on the right basis, and whether the timing of demand or any Section 20B notice still protects recovery. A project can be procedurally well understood at notice stage and still become difficult to defend if the invoice trail is incomplete, late, or impossible to map back to the consulted estimate.

For a block manager or RTM finance lead, this is not just a legal compliance note. It is an operating control. Every invoice needs enough structure around it to show which work package it belongs to, when the cost was incurred, how it will be apportioned, and whether it has already been billed or still needs a demand or notice. If those links are not built while the project is live, the year-end accountant pack turns into a reconstruction exercise instead of a review of an already defensible record.

Build a tender-versus-actual register before the first invoice ages

The register should exist before the first supplier bill lands. If the team waits until year-end, it loses the two details that are hardest to rebuild later: when each cost was actually incurred and how that invoice relates to the consulted estimate pack. A proper major works register is not a generic purchase ledger export. It is a tender-versus-actual control sheet designed around Section 20, Section 20B, and later accountant review.

At minimum, the register should capture:

• contractor name
• invoice number
• invoice date
• incurred date
• gross amount
• VAT amount
• cost category
• work package or budget line
• consulted estimate reference
• apportionment basis
• amount demanded
• Section 20B deadline or notice status
• evidence link back to the source invoice
• exception status

Those fields matter because they answer different questions. Invoice date tells the team what arrived from the supplier. Incurred date supports the recoverability timeline. The consulted estimate reference shows whether the cost belongs to the scope that leaseholders were consulted on. The apportionment basis turns a project cost into a leaseholder charge model. Exception status stops unusual items from disappearing into a total that looks tidy but cannot be defended later.

If the team wants to extract contractor invoices into Excel rather than key those columns by hand, Invoice Data Extraction can take uploaded invoices plus a prompt-defined column list and return structured Excel, CSV, or JSON with fields such as supplier name, invoice number, dates, VAT, totals, and line items. That fits the wider discipline of invoice processing for property management, but the register for Section 20 work should be shaped around consulted estimates, timing controls, and evidence links rather than generic AP coding.

Reconcile each invoice to the consulted scope, contractor, and work package

Each invoice should go through the same control sequence when it arrives. First, confirm the supplier matches the contractor or adviser expected for that part of the project. Next, identify the work package the invoice relates to. Then map it to the closest line in the Statement of Estimates and record any variance in wording, quantity, or price. Only after those checks should the cost move into the recoverable spend trail.

That order matters because "within budget" is not the same as "within consulted scope." A GBP18,000 invoice for roof repairs may still create a problem if the consulted estimate described a narrower package of patch repairs, or if the invoice bundles additional works that were never evident in the estimate pack. The register therefore needs a variance note that captures what changed: broader scope, different contractor, higher quantity, missing breakdown, or unclear supporting papers.

Professional fees also need their own treatment. Surveyor invoices, project-management fees, legal costs, and other advisory spend should not be folded into contractor works lines just because they sit within the same project. They need separate categorisation, separate supporting documents, and often separate review, especially where the lease basis or service-charge treatment is not identical to the works themselves.

The exceptions worth surfacing early are usually the ones the broad Section 20 explainers never get into: an invoice from a supplier not visible in the consulted estimate pack, a work description that has drifted beyond the original scope, a combined invoice that mixes recoverable and non-recoverable elements, or a fee line with no supporting engagement evidence. A reconciliation register earns its value by making those items visible while the job is still live, not by hiding them inside a single project total and hoping year-end review will sort them out.

Track Section 20B deadlines and recoverability before demands go late

Section 20B turns timing into a control field. For each invoice, the register should show when the cost was incurred, the 18-month deadline by which recovery needs to be protected, whether a demand has already been issued, and whether a Section 20B notice has been served if final billing is not yet ready. Without those dates in the live register, the team is left piecing together recoverability from email threads, supplier PDFs, and draft statements after the safest point to act has already passed.

This is also where billing stages need to stay separate. On-account demands, final balancing charges, and later adjustments should not be collapsed into one undifferentiated "charged" column. The reader needs a trail that shows what has already been demanded, what is still provisional, and what still depends on a notice because the final account is not complete. That separation becomes critical when large projects run across accounting periods or when final invoices arrive long after the initial billing round.

A practical deadline tracker does not need to be elaborate. It needs to be visible, dated, and owned. Once an invoice is logged, the next action should be clear: ready to demand, hold pending scope review, hold pending apportionment, or escalate because a Section 20B protection step is needed. The control mindset is similar to the dated-evidence discipline used in UK letting agent client account reconciliation, even though the risk here is recoverability through the service charge rather than a cashbook mismatch.

The dangerous cases are the quiet ones: invoices approved operationally but not pushed through to billing, delayed professional fees sitting outside the main works file, or records that cannot prove when the cost was incurred. Those are not admin nuisances. They are the entries most likely to weaken the eventual recovery position if they are discovered only after the project has moved on.

Apply lease apportionments and isolate the exceptions that distort the final charge

Once an invoice has been reconciled to scope and timing, it still is not ready for leaseholder billing until the apportionment basis is clear. The register needs a field that points to the relevant schedule or lease basis for that cost pool, whether the charge is split by fixed percentage, floor area, block, or another lease mechanism. Without that link, the finance team is not running a service-charge model. It is only collecting supplier spend.

This is why works costs, professional fees, reserve-fund offsets, insurance recoveries, and other adjustments should stay visibly separate. A final contractor invoice may be fully recoverable through the major works charge, while a professional fee needs closer support and a reserve contribution needs different presentation in the balancing position. If those categories are rolled together too early, the later leaseholder calculation looks neat on paper but becomes hard to justify line by line.

Variance also needs to stay in view. When actual invoices materially exceed the Statement of Estimates, the right response is not to bury the overrun in a later balancing charge and hope the project narrative explains it. The register should show the consulted amount, the actual amount, the delta, and the reason for that delta, such as additional quantities, changed specification, delayed professional input, or an omitted line in the estimate pack.

The same approach helps with disputed or unusual items. If a cost category looks marginal, the lease wording is unclear, or an item needs management judgement before recovery, it should remain flagged with supporting papers attached to the register entry. That keeps the apportionment model honest and prevents the year-end file from inheriting a set of unexplained charges that no one wants to defend later.

Prepare the accountant pack so year-end review does not rebuild the project from scratch

By year-end, the major works file should already read like an accountant pack. That means a live invoice register with source references, the consulted estimate mapping, the apportionment schedules used for billing, supporting invoices, and a record of what was demanded or protected under Section 20B. If any of those pieces sit in separate inboxes or personal spreadsheets, the final review starts by rebuilding the project rather than testing a controlled record.

The real payoff is audit trail quality. A clean tender-versus-actual register lets the accountant see how consulted estimates turned into actual invoices, how those invoices turned into leaseholder charges, and where any balancing adjustments came from. That is what makes the file easier to carry into a TECH 03/11-style review instead of rebuilding support from scratch. For broader close-out context, the same control logic carries into UK residential service charge year-end reconciliation, but the advantage here is that the major works evidence has already been structured before year-end pressure arrives.

This is where prompt-defined extraction can help without drifting into legal overclaim. Invoice Data Extraction can convert invoices into structured output and include a reference to the source file and page number on each row, which makes it easier to reuse the same invoice data and document links in the accountant pack. That supports the evidence chain. It does not replace lease analysis, statutory consultation, or judgement on recoverability.

The strongest process captures invoice data once, keeps every exception attached to the relevant row, and carries that record from live project control into the final service-charge review. When that happens, the accountant is checking a coherent file rather than reconstructing a statutory project from scattered PDFs.