For procurement and legal review

Data sourcing and compliance

What a review team needs to know before approving a dataset: where the data comes from, how it is gathered, what personal information it contains and what is delivered with an order.

How data is sourced

Direct Dataset is an independent provider of public data. Catalogue datasets are compiled from publicly available information: public records, published catalogues, advertised listings and published prices and reviews.

  • Each dataset page names its source and carries a source and method statement.
  • Each dataset page publishes its coverage, data dictionary and field definitions before purchase.
  • Direct Dataset is not affiliated with, endorsed by or sponsored by the organisations or platforms whose public information its datasets describe.

The full process is on the methodology page.

Personal information

The datasets describe properties, products, prices and businesses. They are not built to profile individuals.

  • Where a source publishes business contact details or public display names, the dataset page says so in its data dictionary.
  • A version of any dataset without those fields can be delivered.

Your use of the data

The licence in your quotation sets out what you may do with a dataset. You remain responsible for making sure your own use meets the laws and internal policies that apply to your organisation. If your review needs something specific in writing, ask before you order and it can be addressed in the quotation.

Licence types are summarised on the licensing page.

Due diligence questions

Where does the data come from?

From publicly available information: public records, published catalogues, advertised listings and published prices and reviews. Each dataset page names its source.

Is Direct Dataset affiliated with the sources?

No. Direct Dataset is an independent provider. It is not affiliated with, endorsed by or sponsored by the organisations or platforms whose public information its datasets describe.

How is the data gathered?

Each source is recorded on a schedule, then standardised into flat tables with stable identifiers. For sources that change daily, each change is kept as a dated event. The methodology page describes this in full.

Do the datasets contain personal information?

The datasets describe properties, products, prices and businesses. Some include business contact details that a business has published about itself, or public display names on reviews. The dataset page says so wherever that is the case, and a version without those fields can be delivered.

What documentation comes with an order?

A data dictionary, the record count for the agreed scope, the source and method statement and the definitions behind each field. The same information is published on the dataset page before purchase.

What licence terms apply?

The quotation states the licence. Internal use is standard. Research, publication and redistribution terms are summarised on the licensing page.

How are files delivered?

By a time-limited download link, as Parquet, CSV or SQLite. Parquet files load directly into Snowflake, BigQuery, Databricks, DuckDB and pandas.

Can we evaluate the data before buying?

Yes. A sample in the delivery format is available for every dataset, cut to the areas, categories or dates you want to test.

How do we know a figure is accurate?

Every count on this site is measured from the data. The quotation states the exact count for your scope, and a record count by period can be supplied so coverage can be reconciled against figures you already hold.

Questions not answered here

Send your vendor questionnaire or your questions to hello@directdataset.com.