UK Supplier Payment Times: Fastest and Slowest Reports

Compare the same reporting period

We analysed the GOV.UK payment-practices export downloaded on 9 October 2026. The comparison uses reports covering exactly 1 January to 30 June 2026. Mixing each business's latest filing would put different trading periods into the same table.

After removing superseded, incomplete and inconsistent records, the cohort contains 3,227 reports from 3,227 legal entities. It covers businesses that filed for this particular period and passed the checks below. Businesses with different reporting calendars, missing filings or unusable data are outside the comparison.

Download the complete 3,227-entity results table (CSV). It includes company numbers, filing dates, report links, payment-speed bands and the separate late-payment percentage. Company numbers should be imported as text to preserve leading zeros.

The shortest reported averages need a careful reading

Two entities reported a rounded average of zero days. Both reported that 100% of payments made were within 30 days and that 0% of payments due missed agreed terms.

These are the lowest values in the filtered January–June 2026 cohort, tied on reported payment speed. The figures are whole-number averages. Zero therefore does not establish that every payment was immediate.

Mott MacDonald Bentley's explanation is particularly important. The entity says it makes payments only to joint-venture partners and none to suppliers outside the Mott MacDonald group. Its zero-day figure tells an outside supplier little about how their own invoice would be handled. Do not transfer this result to another similarly named group company.

Pioneer Shipping says its most frequently used terms are 10 business days after the date or receipt of an invoice for vessel charter hire. Those contractual terms and its reported zero-day average answer different questions. We retained the filed figure, rather than treating the terms as evidence that the average must be wrong.

Nine further entities reported a one-day average. Even that apparently simple comparison can conceal a different payment mix. Mercury Pharmaceuticals Limited's filing reports one day overall but says the average for external vendors, excluding intercompany vendors, is 41 days. Both are the company's reported figures. We have no underlying invoice records with which to audit them.

The longest reported average payment times

The five longest averages in the same cohort are below. The final column measures a different population: payments contractually due in the period that were not paid within agreed terms.

All five reports cover January–June 2026. These are reported legal-entity results, not whole-group scores or independently verified findings about misconduct. The linked filings contain their payment terms and explanations.

Admiral Law includes professional disbursements for legal cases, such as counsel fees, many on extended credit terms. It reports an alternative average of 48 days when professional disbursement invoices are excluded. That narrower figure is useful context, but substituting it into the ranking would compare different populations.

Ernst & Young (Asia-Pacific) Services includes both external suppliers and contracts with firms in the wider EY network. It reports an external-supplier-only average of 122 days, compared with 547 days overall. This result belongs to the named UK legal entity. It cannot describe every EY firm or country.

Fresenius Medical Care Renal Services says it pays many intercompany invoices and that these affect the reported average. Its filing describes 180-day group-supplier terms, compared with predominantly 60-day third-party terms. It gives no separate external-only average that we can put beside the 249 days.

IBC Vehicles and Petrofac Engineering describe payment terms in their filings, but do not provide a separate external-only average. We cannot identify the cause of their long averages from the published statistics. A report can include payments made this half-year against invoices received much earlier, so an average longer than six months is possible.

What the middle of the cohort looks like

Each entity counts once in these results, regardless of how many payments it made.

Source: StandardsDesk calculations from the official export, snapshot 9 October 2026. Each included entity counts once. Percentages in the last two rows are reported payment metrics, while 47.0% and 9.1% are shares of entities. The 30-day, 60-day, 25% and 10% cut-offs are descriptive choices for this comparison, not regulatory ratings.

The 325 fast-average but frequently late reports show why speed alone is insufficient. A business can pay fairly quickly and still miss short agreed terms. Conversely, long terms can produce slow payment with relatively little contractual lateness.

The 32-day median means the middle entity's reported average was 32 days. It does not mean half of all UK invoices were paid within 32 days. The public export does not provide the payment counts needed to combine these entity averages into an invoice-weighted national average.

Payment speed and lateness use different denominators

Under the government reporting guidance, the average time to pay is a mean for payments made under qualifying contracts during the reporting period. It runs from receipt of an invoice or other payment notice to payment, subject to the guidance's specific rules. Invoices still unpaid are excluded from that average.

The late-payment percentage instead concerns payments contractually due during the period that were not made within agreed terms. It can include unpaid and disputed invoices. Invoices already overdue at the start of the period are not counted again in that measure.

Consequently, the average, the over-60-day share and the late-payment percentage cannot be subtracted from one another. Admiral Law's 94% paid after 60 days and 4% outside agreed terms are not contradictory by themselves. The timing thresholds and contractual deadlines differ, as do the populations counted.

Intercompany payments also belong in the reporting when they arise under qualifying contracts between separate businesses. An external-supplier ranking would require a consistent dataset. Most entities do not publish the separate figures needed to build one.

How we selected and checked the reports

The unedited export contained 115,427 historical records. Of these, 3,553 covered exactly 1 January–30 June 2026, representing 3,529 company numbers before cleaning.

We used company registration numbers to identify legal entities. When an entity had a newer filing covering the same dates or overlapping by more than 31 days, we retained the newer filing. For same-day filings, the higher report ID takes precedence. This removed 24 superseded records. No newer off-calendar overlapping report displaced an otherwise included exact-period record in this snapshot.

We then removed 268 reports indicating no qualifying contracts or no payments, 33 with inconsistent invoice values, and one with missing or non-integer metrics. The remaining 3,227 records each represent a different entity.

The value checks require a positive total across the three paid-invoice value bands and a non-negative reported late-payment value no larger than that total. Count percentages must fall between zero and 100, and the three payment-speed bands must sum to 99–101% to allow whole-number rounding. Average days must be non-negative. We also checked for premature filings, although that rule removed no records here.

These checks draw on the Department for Business and Trade's July 2026 methodology. Our comparison is narrower than its annual statistics: one exact half-year, additional count-band checks, and entity rankings. Requiring coherent value fields also means a report with usable count figures can still be excluded. We have kept that rule explicit rather than treating every filed number as usable.

Equal day values share a rank. Alphabetical order within a tie does not imply better performance. We retained zero-day and unusually high averages, checked the named reports against their live company reporting histories, and read the explanatory text before presenting them. A separate calculation reproduced the selected report IDs, medians and extreme values.

The register can include voluntary reports. We did not independently retest each entity's size eligibility. Businesses that do not file are not represented, and the results are not adjusted for them.

What a supplier should check before agreeing credit

Start with the company number on the contract or purchase order. A familiar brand may have several reporting entities. Search that specific business and check the period dates, not just the filing date.

Read its standard terms and notes alongside average days, payment-speed bands and lateness. Look for intercompany transactions, client disbursements, special supplier terms and changes in the reported payment mix. Compare successive equivalent periods if you need a trend. This single-half-year study cannot establish one.

Then put the agreed collection date, and a plausible delayed date, into your cash forecast. Our cash-buffer study shows how a delayed receipt changes the cash a business needs. The UK late-payment cost analysis separates an unpaid balance from the expense of funding it.

If borrowing is part of the contingency, check the published UK lender requirements before relying on finance being available. You can use the business loan calculator to test a stated rate, fee and repayment schedule. Those repayments belong in the forecast too.

Frequently asked questions

Can I treat a missing report as a good or bad payment record?

No. This analysis does not establish why an entity is absent. It may have a different reporting calendar, be outside scope, not have filed, or have a record excluded by our checks. Absence does not give it a score.

Are the figures independently audited?

The businesses submit the reports. DBT's methodology says the reporting data are not independently verified or audited. Our checks identify some inconsistencies and preserve the reported explanations, but do not verify the underlying payments. Later corrected or backdated filings can change the comparison.