How the figures work
Methodology and limitations
Waiting-time calculation
For a report marked paid, waiting time is the number of calendar days from submission to payment. For every unresolved status, it is the number of calendar days from submission to today, so it continues to increase. The server calculates this value; the browser does not submit it.
Primary measure
We lead with the median: half of reported waits are shorter and half are longer. The average is also shown, but a few unusually long waits can distort it. The 25th and 75th percentiles describe the middle half of reports. Minimum and maximum figures may be unrepresentative.
Open and completed reports
A report is treated as completed only when the contributor records a payment date. Every other report is labelled open, including one that has been approved but not yet marked paid. For an open report, the displayed duration is elapsed time so far and increases each day.
The headline median, average, percentiles and charts combine open elapsed-so-far durations with completed payment waits. They describe the current submitted dataset and must not be interpreted as the typical time taken to complete a VAT repayment. The number of open and paid reports is shown beside the figures so this limitation is visible.
Status history and approval
Contributors can use their private report token to add a dated status change. The service retains the earlier event instead of replacing it. Time to approval is calculated from the original submission date to the date the contributor records the repayment as approved or confirmed.
Inclusion and small samples
Valid reports are published automatically and enter the public figures. Flagged and rejected reports are excluded. Statistics are shown at all sample sizes, including when only one report matches. A small-sample aggregate can reveal the values contributed by one person and should not be treated as anonymised or representative.
Data quality
Reports are self-submitted and are not independently verified. Validation rejects impossible date combinations, future dates and claims outside the configured 24-month window. Duplicate detection and submission controls reduce abuse, but cannot eliminate it.
Limitations
This service has only recently launched, so the available dataset is currently limited. Users of this site are a self-selected sample and may not represent all VAT repayments. Results describe this sample only and cannot predict when any individual repayment will arrive.