How Buildings Catch Billing Errors Before They Are Paid

Abstract#

A billing error is cheapest to fix before the invoice is paid. This blog tests one question: which errors on a commercial waste invoice can a building catch before payment and what record does each check need. It draws on the Department for Business and Trade statistics on how quickly large businesses pay and on a public sector audit programme for accounts payable. It reviews the duty of care code of practice and the Digital Waste Tracking (England) Regulations 2026. It sets them beside three peer reviewed studies located through Google Scholar and published material from Wastify AI. The finding is that large UK businesses paid suppliers in a median of 32 days in 2025 and that the standard control for that window is a three way match of invoice and order and receiving record. For waste the receiving record is usually missing. Of nine common error types a building holding only the contract and the invoice can test three before payment and one more in part. The other five need a collection record held by the building. The economics of credence goods explains why: when a buyer cannot verify what was supplied the information gap sits with the buyer.

Key Findings#

1. Large UK businesses paid their suppliers in a median of 32 days in 2025 and 15% of invoices by number were paid late. The figures come from 11,178 valid reports [2]. That window is the only point at which a waste invoice can be checked before money leaves the building.
2. The standard accounts payable control is a three way match between invoice and purchase order and receiving report. A public sector audit programme asks auditors to verify that goods or services were received and to pay only for the portion received and to look for duplicate payments [3]. For waste the receiving report is usually missing.
3. Of nine common waste invoice error types a building with only the contract and the invoice can test three before payment and one in part: 33% in full. Five need a collection record held by the building [3 4 5].
4. Credence goods are services whose buyers worry about paying for work not provided and where that worry does not disappear even after the bill is paid [7]. In a field experiment buyers who appeared not to know the local tariff system faced a likelihood of a manipulated bill about fifteen percentage points higher [8].
5. A review of every published econometric study since 1970 found that cost savings from contracting out waste services are not systematic. Transaction cost theory argues that contracting out works best where contracts are complete: a rare situation in public service markets [9]. Wastify AI states that buildings that start verifying collections routinely find discrepancies in the first month [1].

1. Introduction#

1.1 The Question#

A waste invoice arrives weeks after the collections it describes. It lists lifts and containers and charges. In most buildings it is approved because it resembles last month and because the building holds nothing to test it against. Once paid any error becomes a recovery exercise in a closed period.

This blog asks one question. Which errors on a commercial waste invoice can a building catch before payment and what record does each check need. The question can be tested against standard accounts payable practice. If the documents a building already holds were enough to test every error type there would be no gap. Section 3 shows that five of nine error types need a record most buildings do not keep.

1.2 Why This Matters Now#

Two developments change what a building can check and what it is expected to check.

First receivers now record weight and containers for every load. Under the Digital Waste Tracking (England) Regulations 2026 a receiver must record the total weight under each waste code with the unit of measurement and whether the weight is estimated or actual. It must also record the type and number of containers [5]. A second record of each collection now exists outside the invoice. It is held by the receiver rather than the building.

Second the test of who may carry waste is changing. Research Guide 09 in the wider Wastify AI research programme found that from 22 July 2027 controlling or transporting waste in England will need an environmental permit or a registered exemption under the carrier permitting reform unless an exclusion applies. Until then the check is a waste carrier registration and the impact assessment records that on application there are no initial checks of the basic information provided [10]. Confirming who is being paid is becoming part of confirming what is being paid for.

1.3 Building on the Research Series#

This is the seventh blog in the Wastify AI research series [6].

• Research Blog 03 Found that the independent review of service charge accounts checks for duplicate transactions but excludes the allocation and apportionment of the service charge.
• Research Blog 06 Found that the published rules allow one lift of a 1,100 litre container to be estimated at values 7.1 times apart.

Both findings point to the same moment. The accountant review comes after payment and the estimate is built before it. The approval of each invoice is where a building record can do its work.

1.4 Scope and Structure#

This blog covers invoices for waste collection services at multi tenant commercial buildings in England. It proceeds in six sections.

• Section 2 Describes the method and the sources.
• Section 3 Sets out five findings.
• Section 4 Discusses what they mean for property managers and finance teams and contractors.
• Section 5 States the limitations.
• Section 6 Answers the question.

It makes no claim about the billing practice of any contractor. Level 0 is the competition: the gap is the absence of a building record rather than the conduct of any supplier.

Test the invoice while it can still be corrected. Wastify AI delivers waste intelligence for commercial buildings where every collection is weighed and confirmed by stream before it leaves.

2. Method#

2.1 Research Design#

The blog combines three evidence streams. The first is public data and guidance on the payment window and on standard invoice controls: the Department for Business and Trade payment statistics and a public sector audit programme for accounts payable together with the duty of care code and the Digital Waste Tracking Regulations. Nine common waste invoice error types were then mapped against the record each check needs. The second stream is peer reviewed research on markets where buyers cannot verify what they receive and on contracting out waste services. Three studies were located through Google Scholar and each abstract was read at a recognised index. The third is published material from Wastify AI.

2.2 Sources#

External sources were opened directly between 7 and 13 October 2026 and each reference carries its access date.

Table 1: Sources consulted

Source: compiled for this blog from the references listed at the end.

2.3 Platform Data#

Contractor Mode features and the statement on discrepancies are taken from the published Contractor Mode and contractor verification pages. The container register holds tare weights for eleven bin types from 2.5 kg to 58 kg. Platform totals were read from the live counter on the Wastify AI homepage on 13 October 2026. The counter showed 5,149,000+ kg weighed at disposal and 958,000+ verified weighings and 380+ companies reporting. The same page lists £660,000+ recharged on verified data. Public totals are rounded down.

2.4 Calculations#

Three of nine error types testable in full gives 33.3% and is stated as 33%. Four of nine testable in full or in part gives 44.4%. Five of nine needing a building record gives 55.6%. Every derived figure was recomputed from the stated inputs before publication.

3. Findings#

3.1 The Window Before Payment#

The Department for Business and Trade publishes statistics from the payment practices reports of large businesses. In 2025 large businesses paid their suppliers in 32 days. Fifteen per cent of invoices by number were paid late. The statistics use median values and rest on 11,178 valid reports [2].

Thirty two days is the window. Before it closes a disputed line can be held and queried while the collection is recent and the people involved remember it. After it closes the error moves into the service charge and then into the year end accounts. Research Blog 03 found that the accountant review of those accounts checks for duplicate transactions and reconciles the statement to the accounting records but excludes the operational review [6]. The review comes too late and looks at the wrong things to catch a quantity error.

3.2 The Three Way Match and the Missing Receipt#

Accounts payable teams in most sectors rely on a three way match. A public sector audit programme from the Office of the State Comptroller in New York sets out the control. A three way match is performed between the invoice and the purchase order and the receiving report. The auditor determines whether the goods or services purchased have been received and verifies that a partly filled order is paid only for the portion received. Payment history is reviewed for duplicate payments [3]. Table 2 sets out what each document becomes for a waste invoice.

Table 2: The three way match for a waste invoice

Source: audit programme for accounts payable [3] and duty of care code of practice [4].

The order exists: it is the contract schedule of containers and rates. The invoice exists. The receiving report usually does not. The duty of care code requires a transfer note recording the classification of the waste and its quantity and nature and whether it is loose or in a container [4]. A quantity expressed as a number of containers meets that requirement and a single season ticket can cover a series of transfers of the same waste with the same carrier for up to one year [4]. A building that cannot show what left on a given day can only run a two way match.

3.3 Nine Errors and the Record Each Needs#

Table 3 lists nine common error types on a commercial waste invoice and the record a building would need to test each one before payment.

Table 3: Nine waste invoice error types and the record each check needs

Source: framework for this blog drawing on [3] [4] [5] and the Wastify AI record structure [1].

The pattern is stark. With the contract and the invoice alone a building can test three error types in full: duplicate lines and rates that differ from the schedule and arithmetic. It can test unrequested ad hoc lines in part if it keeps a log of requests. Five error types need a record of each collection held by the building: lifts charged that did not happen and the wrong container count or size and the wrong stream rate and weight based surcharges and contamination charges without evidence. Those five are the errors that turn on what actually left the building.

3.3.1 Hazardous Waste Already Has a Receipt#

One category of waste shows what a full receiving record looks like. A hazardous waste consignment note must accompany every movement of hazardous waste in England and Wales. The receiving site then makes a consignee return each quarter to the Environment Agency and separately to the producer. For those loads the building holds a document for each movement and receives the figures back from the site. For general and recycling streams there is no equivalent. Electrical waste collected under the WEEE regulations is often billed as a separate ad hoc line and needs the same evidence of what was collected.

3.4 Why Information Protects the Buyer#

Economists call a service a credence good when the buyer cannot judge what was supplied. The review of the economics of credence goods by Dulleck and Kerschbamer begins with the buyer who worries about paying for a service that was not provided. It notes that this concern does not disappear even after the repair and the payment of the bill [7]. The review asks which institutions and market conditions solve the problem. Verification by the buyer is the one this blog examines.

A field experiment on credence goods in Athens measured how information changes outcomes. Passengers with inferior information about optimal routes were taken on significantly longer detours. Passengers who appeared not to know the local tariff system faced a likelihood of a manipulated bill about fifteen percentage points higher [8]. The study concerns taxis. Its finding is general: the buyer who visibly holds the information is the buyer who is billed correctly.

The review of waste service contracting adds the contract side. Bel and Warner reviewed every published econometric study of water and waste production since 1970 and found that cost savings are not systematic in waste. They note that transaction cost theory argues contracting out is best where contracts are complete: a rare situation in public service markets [9]. A waste contract priced per lift on an assumed fill is incomplete by design because the quantity it pays for is never observed by the buyer. Wastify AI puts the same point plainly on its contractor verification page: waste collection is one of the few services where the supplier provides both the service and the data that prices it [1].

Table 4: Peer reviewed studies used in this blog

Source: published abstracts of the three studies [7 8 9].

3.5 What a Building Record Adds#

A collection record closes the five gaps in Table 3. With collection weighing under a contractor PIN the contractor enters a unique PIN on the kiosk in the waste room and each bin is weighed and confirmed by stream and timestamped against the building before it leaves [1]. The contractor invoice verification by weight then checks the invoice line by line against those weights and flags discrepancies with the evidence attached [1]. Wastify AI states that buildings that start verifying routinely find discrepancies in the first month [1].

The record is net of the container through the tare weights for eleven bin types which matters for any weight based surcharge [1]. Across the platform 958,000+ weighings covering 5,149,000+ kg have been recorded [1]. The verified weights then carry into the monthly cost ledger by component where each occupier share is produced from recorded costs and verified weights and every line traces back to a weighed and photographed record [1]. The carrier check sits alongside: confirm the upper tier registration today and the permit or registered exemption from July 2027 [10].

The invoice states what was collected. A building record shows it. Wastify AI weighs every collection so each line can be approved or held on evidence.

4. Discussion: What This Means for the Reader#

4.1 For Property Managers#

The practical rule follows from Table 3. Approve the lines that reconcile and hold the single line that does not. With a collection record the query carries a date and a weight and a stream rather than a suspicion. Without one the building can still run the three checks that need only the contract and the invoice: duplicates and rates and arithmetic. Those three should never be skipped because they cost nothing to run.

4.2 For Finance Teams and Managing Agents#

Finance teams already know the three way match. The waste line is often the one category where it cannot be run because no receiving record exists. The RICS standard requires managers to seek to recover no more than 100% of the proper and actual costs of the provision or supply of the services [11]. A cost that was never received is not an actual cost and the only way to show the difference is a record of what was received. Every allocated line can then be followed through the evidence behind every allocated line to the photograph taken at the scale [1].

4.3 For Contractors#

A building record helps a contractor too. Disputes settle on the same evidence for both parties and correct invoices are approved faster. The research on credence goods concerns information rather than intent and nothing in this blog suggests that any waste contractor bills incorrectly. The incomplete contract described by Bel and Warner is a feature of the market [9]. A shared record makes the contract complete in the one dimension that matters most: what left the building.

5. Limitations#

This blog has five limitations that a careful reader should weigh.

1. The nine error types are a framework built for this blog from standard accounts payable controls and the records required for waste. No regulator publishes this list or measures how often each error occurs.
2. The audit programme cited is from the New York State public sector. The three way match it describes is standard practice in the UK as well but the document itself is not a UK source.
3. The research on credence goods comes from taxi markets and expert services. It is used to explain why information matters and not to suggest that waste contractors behave in the same way.
4. The statement that buildings routinely find discrepancies in the first month is published by Wastify AI. It is not an independent study and no rate of discrepancy is published.
5. The payment statistics describe large businesses in all sectors and the reporting thresholds changed in 2025. Waste invoices may be paid on different terms.

6. Conclusion#

A building has a window of about a month to catch a waste billing error before it is paid. The standard control for that window is a three way match and for waste the receiving record is usually missing. With the contract and the invoice alone a building can test three of nine common error types. The five that turn on what actually left the building need a collection record that the building holds itself. Approval without a receipt is acceptance rather than control.

A building that wants to know which of the nine checks it can run today can start with a check of its waste data in two minutes.

Approve your next waste invoice against your own record. Book a demo with Wastify AI for a one hour session tailored to your buildings.

Frequently Asked Questions#

What are the most common waste billing errors?#

Duplicate lines and rates that differ from the contract and arithmetic errors. Lifts charged that did not happen and the wrong container count or size. The wrong stream rate and weight based surcharges and contamination charges without evidence and unrequested ad hoc lines.

Which errors can a building catch without its own collection record?#

Three in full: duplicates and rate differences and arithmetic. Unrequested ad hoc lines can be tested in part if requests are logged. The other five need a record of each collection [3 4].

What is a three way match for waste?#

The invoice matched to the contract schedule and to a record of what was actually collected. The first two documents usually exist. The third usually does not unless the building weighs its own collections [3].

How long does a building have to check an invoice?#

Large UK businesses paid suppliers in a median of 32 days in 2025 [2]. That is roughly the window for a query before payment.

Does a waste transfer note prove what was collected?#

Not on its own. It records the quantity and nature of the waste and a number of containers meets the requirement. A season ticket can cover a series of transfers for up to one year [4].

Should the whole invoice be held over one disputed line?#

No. Approve what reconciles and hold the single line with its evidence. That keeps the relationship with the contractor straightforward.

Does checking invoices mean distrusting the contractor?#

No. The research concerns information rather than intent. A shared record settles disputes faster for both parties and makes correct invoices quicker to approve.

Can a building ask questions of its collection records?#

Yes. Living AI answers plain English questions about the weighings with the sources attached so a question about the collections on a given day is answered from the records.

References#

[1] Wastify AI published material and platform data. Contractor Mode and contractor verification pages including the statement that buildings that start verifying routinely find discrepancies in the first month. Container register of eleven bin types from 2.5 kg to 58 kg. Platform totals of 5,149,000+ kg and 958,000+ verified weighings and 380+ companies reporting read from the homepage counter on 13 October 2026. £660,000+ recharged on verified data as listed on the same page. wastify.co.uk.

[2] Department for Business and Trade. Large businesses payment practices and performance statistics 2025: commentary. Published 14 July 2026. gov.uk. Accessed 13 October 2026.

[3] Office of the State Comptroller New York. Internal controls over accounts payable: audit programme. osc.ny.gov. Accessed 13 October 2026.

[4] GOV.UK. Waste duty of care code of practice. Sections 3.5 and 3.6. gov.uk. Accessed 7 October 2026.

[5] legislation.gov.uk. The Digital Waste Tracking (England) Regulations 2026: Schedule 1. SI 2026/729 paragraph 14. legislation.gov.uk. Accessed 12 October 2026.

[6] Wastify AI Research Blogs 01 to 06. October 2026. wastify.co.uk.

[7] Dulleck U and Kerschbamer R. On doctors and mechanics and computer specialists: the economics of credence goods. Journal of Economic Literature 44(1) pages 5 to 42. 2006. ideas.repec.org. Accessed 13 October 2026.

[8] Balafoutas L and Beck A and Kerschbamer R and Sutter M. What drives taxi drivers: a field experiment on fraud in a market for credence goods. Review of Economic Studies 80(3) pages 876 to 891. 2013. ideas.repec.org. Accessed 13 October 2026.

[9] Bel G and Warner M. Does privatization of solid waste and water services reduce costs: a review of empirical studies. Resources Conservation and Recycling 52(12) pages 1337 to 1348. 2008. ideas.repec.org. Accessed 13 October 2026.

[10] Wastify AI Research Guide 09. From Registration to Permit: Waste Carrier Reform and Due Diligence for Building Managers. Document reference CAI/GR/009/WAI. October 2026. wastify.co.uk.

[11] RICS. Service charges in commercial property. Professional standard UK 2nd edition. Published June 2025 and effective 31 December 2025. Requirement 2. rics.org. Accessed 9 October 2026.