How to evaluate data freshness

Businesses open, close, move and change numbers every day, so any business dataset starts ageing the moment it is produced. Claims such as "freshest data online" are impossible to verify. Look for evidence instead.

Evidence worth asking for

• Last processed date. When the dataset was last compiled or cleaned. Every product page here shows it, taken from the processed files themselves.
• Version history. A dated changelog shows how often the data is actually refreshed.
• Field coverage. A dataset where 99% of rows have a phone number but only 20% have a website tells you something about its source.
• Structure preview. You should be able to see the exact columns before paying.

What we show

On each dataset page: version code (e.g. UK-BIZ-2026-09), last processed month, changelog and measured coverage for each field.

Testing freshness yourself

Buy the trial or a single pack, take a random sample of 50 rows and check them against the businesses' own websites or public listings. Count how many still match. A sample of 50 gives a rough indication, not a precise rate — the margin of error at that size is roughly ±14 percentage points.