How we build, source and refresh every figure on the Data Hub.
Each metric across the five hubs is bound to a named primary source, carries the observation period and last-checked date, and is refreshed on a 30-day cycle against the source. This page documents the standards, the source ladder, the caveats we apply, and how to cite us.
What we promise on every figure
Four standards apply to every metric published on this hub.
Primary source or none
Every figure is bound to an identifiable primary source — a regulator, a national statistical office, a trade body’s published return, or a scheme operator. Aggregator sites, blog posts and consultancy press releases are never used as the source of record.
Period stated, always
Every figure carries the observation period it covers (e.g. “Q1 2026”, “CY 2024”, “Jan 2019 → Mar 2026”). We do not publish a number without a period; we do not roll forward stale numbers.
Refreshed on a 30-day cycle
We refresh-check each metric against its source every 30 days. The “last checked” date appears next to the figure. If a source has updated and our number hasn’t, the audit log flags it before the page goes stale.
Mandatory caveats stay attached
Where a source carries a known structural caveat — UK Finance’s 97% coverage, ONS’s Visa-only spending data, the BNPL hidden-debt gap — that caveat travels with the figure and appears on every page that uses it.
How we choose between sources
When two or more sources publish the same metric, we use a fixed precedence order rather than picking a number that sounds right.
| Tier | Source type | Examples | When we use it | Status |
|---|---|---|---|---|
| 1 | Government / regulator | Bank of England · ONS · FCA · PSR · CMA · HM Treasury · NCA | Always preferred where they publish the metric. Used as the source of record. | Source of record |
| 2 | Public-interest industry body | UK Finance · BRC · OBIE · Pay.UK · UK Cards Association | Where members file structured returns and the body publishes aggregate figures with a stated coverage rate. | Source of record |
| 3 | Scheme / network operator | Visa · Mastercard public schedules · Pay.UK Faster Payments stats | For scheme-specific data only (interchange schedules, scheme volumes). | Scheme-specific only |
| 4 | Industry survey | BRC Payments Survey · FCA Financial Lives Survey | Where no direct measurement exists. Survey-derived flag attached. | Survey-derived |
| 5 | Aggregators / consultancies / press | — | Never used as source of record. May appear in linked context, never bound to a metric. | Excluded |
UK Finance and PSR both publish APP fraud figures. UK Finance fraud reports are tier 2 (industry body, structured member returns); the PSR’s APP-scams Performance Report is tier 1 (regulator). Where both cover the same period, we use the PSR figure and note where UK Finance differs.
How a metric goes from source to page
Every metric on the hub follows a defined six-step lifecycle. Each step is timestamped and the audit log is exported to the refresh-audit page.
Source check
Automated daily check of source release schedules. New publications trigger a refresh task.
Bind
Metric is bound to a named source — a publication, a release version, a table reference. The bind is permanent for the metric’s life.
Verify
Two-person check: one analyst pulls the figure, a second analyst independently re-pulls and matches.
Caveat
Mandatory caveats are attached. Non-comparable changes (definition shifts, regulatory transitions) are flagged.
Publish
Metric appears on the page with its period, source, and last-checked date. The audit log entry is written.
Re-check (30d)
Every 30 days, the bound source is re-checked. Stale metrics are flagged on the page until refreshed.
The five caveats that travel with every relevant figure
Some structural caveats apply across the hub. They are attached to every figure they touch, on every page they appear on, in their full form.
1. UK Finance coverage (≈ 97%)
UK Finance returns cover firms representing approximately 97% of UK card issuance and acquiring volume. Smaller issuers, acquirers and PSPs outside membership are not represented. Where a figure is sourced from UK Finance, the 97% coverage caveat applies.
2. BNPL hidden debt
Buy-now-pay-later (BNPL) volumes and balances do not appear in the Bank of England’s Money & Credit dataset and are not in UK Finance card data. Total UK consumer credit exposure cannot therefore be read from these series alone. BNPL is in regulatory transition during 2026; pre-regime and post-regime measurements may not be directly comparable.
3. Effective vs advertised APR
The Bank of England’s effective interest rate measures the weighted-average APR borrowers actually pay on interest-charging balances. Advertised “representative” APR is a different construct, applied only to applicants accepted at that rate. The two are not directly comparable. Where the effective APR figure appears, the gap to representative APR is stated.
4. Visa-network spending data
ONS uses anonymised Visa MCG data for household spending-category breakdowns. This represents the Visa network only and does not capture Mastercard, Amex or scheme-agnostic patterns. Where this dataset is used, the Visa-only caveat applies.
5. SME-list-rates only (card processing)
Enterprise card-processing agreements are privately negotiated. Benchmark tables on the card-processing hub cover SME list and starting rates only. Enterprise pricing is explicitly excluded; we do not publish numbers for it.
Terms used across the hub
These are the operational definitions we use. Where a source uses a different definition, the source’s definition prevails and is flagged.
| Term | Definition | Source |
|---|---|---|
| APP fraud | Authorised push-payment fraud — victim is tricked into authorising a payment to a fraudster’s account. | PSR / UK Finance |
| Effective APR | Weighted-average APR actually paid on interest-charging credit-card balances. | Bank of England |
| Representative APR | The APR at least 51% of accepted applicants are offered at, by definition. | FCA Handbook |
| Persistent debt | FCA threshold: a credit-card account paying more in interest, fees and charges than principal repaid over an 18-month period. | FCA |
| Online retail | Internet sales by retailers, excluding fuel. | ONS Retail Sales Index |
| Mobile commerce | Share of online retail value initiated on a mobile device. | UK Finance |
| CNP fraud rate | Card-not-present fraud value, expressed per £1,000 of online turnover. | UK Finance |
| Effective rate (acquiring) | All-in cost to a merchant of accepting cards, as a percentage of card sales (interchange + scheme + acquirer margin). | BRC Payments Survey |
| Acquirer market share | Share of UK card-acquiring volume. | PSR MR22/1.6 |
| Reimbursement rate (APP) | Share of APP fraud loss reimbursed to victims by sending firm under PSR mandatory reimbursement (from Oct 2024). | PSR APP-scams report |
How to cite a Business Expert Data Hub figure
Each metric on the hub is intended to be citable. Use the form below; include the metric name, period, our source attribution, and the URL.
Suggested citation form
Worked example
Where you cite a metric we have produced (e.g. ratio, share, rolling average derived from primary data), please indicate it is derived and link to the methodology page.
How to flag an error and how the data may be reused
Public-interest data: we want it to be used, and we want errors fixed quickly.
Corrections
If you spot an error — a stale figure, a wrong source, a missing caveat — email data@businessexpert.co.uk with the page URL, the metric, and what you believe is wrong. We acknowledge within two working days, action within five, and log the change on the refresh-audit page.
Licence
Figures sourced from UK government and regulators (BoE, ONS, FCA, PSR, CMA, NCA) are republished here under the Open Government Licence v3.0. Editorial commentary, data tables produced by Business Expert (composites, derivations, ratios) and the layout of this hub are © Business Expert; reuse is permitted with attribution per the citation guide above.
Reuse and attribution
You may reuse figures from this hub in academic, journalistic, regulatory and commercial work, provided you (a) cite using the form above, (b) link to the page, (c) honour the mandatory caveats listed in section 04, and (d) do not represent derivations as primary-source figures.
Contact
The data desk is contactable at data@businessexpert.co.uk. Press: press@businessexpert.co.uk. Source-of-record queries should be directed to the named primary source in the first instance; we are happy to help interpret.