Our principles

This page sets out how DataBuss decides what to cover, how it tests the claims behind every listed database, and what happens when a mistake slips through.

How topics get picked

DataBuss began in 2020 as a short list of links kept while comparing company databases for an unrelated piece of research. Those notes grew into write-ups, and the write-ups grew into a running catalogue that now holds over 500 articles spanning company databases across different industries and regions.

Topic selection follows a fixed checklist rather than editorial whim, which is what keeps two reviews of unrelated databases genuinely comparable. Ninety of the entries in the archive are in-depth reviews, reserved for databases substantial enough to warrant testing beyond a short summary.

How facts are checked

A database earns its place in the ratings only after its search functions, filters, and data accuracy have been tested directly, not after a glance at its homepage. Marketing claims made by a database about itself are checked against what its own interface actually returns for a given query.

The testing process covers the same ground each time, which is what makes the resulting rating meaningful:

• running real search queries against the database rather than relying on sample screenshots
• checking filters and search functions for whether they work as advertised
• comparing the results returned against the accuracy the provider claims in its own marketing

Keeping reviews accurate over time

Reviews are not written once and left to age. Whenever a database changes its pricing or its interface, the older write-up covering it gets revisited rather than quietly going out of date on the page.

Readers who spot something wrong, outdated, or missing in a review can flag it by email at reviews@databaseofcompanies.online. The editors read mail on Mondays and Wednesdays, and a correction is made once the point has been checked against the database itself rather than taken on trust.

Independence and disclosure

DataBuss is an editorial project, not a sales channel. It reviews and rates online databases of companies and publishes what it finds; it does not sell access to any of the databases it covers. Rankings are not for sale, and no database pays its way onto the list or up it.

Plain language is favoured over jargon throughout the archive, on the view that jargon tends to hide what a database actually does rather than explain it. The same principle applies to this page: what DataBuss does, and how it checks its own work, is stated here without dressing it up.