Saimaa ringed seal, Rantasalmi, Finland. Photo: Jan Ebr and Ivana Ebrová, CC BY 4.0
Blog ·
Real-time vs stored B2B data. Why bought lists go stale.
Most prospecting tools sell access to a stored snapshot of companies. It was true when it was collected and starts ageing the same day. The alternative is to start from the primary source, the company register, when you search. Here is the honest comparison.
Two ways to build a company list
Stored database. A vendor collects data from websites and bought lists, merges it into one dataset and sells access to it. Everyone who buys gets the same rows. You get large volumes on well-known companies instantly.
Register first. When you search, the list starts from the official company register and each company's own website. New companies show up as soon as they are registered, and closed ones drop out.
Atlas works the second way, with one honest addition: it also keeps a shared company bank of what users' searches have already found, so it does not have to fetch everything again. Atlas's register data is checked against the official registers every night in seven countries (Finland, Sweden, Norway, the UK, France, Estonia and Singapore), and companies that have closed are marked as closed.
How fast does a stored list age?
We measured it in our own data, so we do not need to quote anyone else's estimate. Our Finnish company data was imported once from the register and never rechecked. A year later, 38.6 % of the records were companies that had already closed, and 50 968 of them still had a working domain that made them look alive. The full audit: we found 289,225 dead companies in our own database.
Closed companies are only one part of the ageing. People change jobs and their email and direct line stop working. Companies move, merge and change names. And every company registered after the snapshot is simply missing.
Who a stored database cannot see
Collecting at scale is cheapest on large, English-language companies with a lot of online presence. A company registered last month, a micro company with a one-page site or one that works only in its local language leaves little trace. It is in the register from the first day, but it may never enter a stored database. In June 2026 we measured that 83 % of 10.3 million operating companies had no website.
Side by side
| Stored database | Register first | |
|---|---|---|
| Freshness | Snapshot that ages from collection | Checked against the register |
| New and small companies | Often missing | Visible soon after registration |
| Closed companies | Stay unless someone rechecks | Marked closed by a nightly register check (seven countries) |
| Source of each record | Often unclear | Shown next to the data |
| Volume on large known companies | Instant and very high | Slower, fetched as you search |
This is not "real-time wins everything". If you need a high volume of large, well-known companies quickly and can live with some bounces, a stored database is faster. Starting from the register wins on freshness and on the new, small and local companies.
The GDPR angle
GDPR requires personal data to be accurate and kept up to date. A list that ages every month is in tension with that. Data from a public register and a company's own website, checked when you use it, is easier to keep accurate. You are still responsible for your legal basis and for how you contact people, whatever the source.
A simple test
Ask any tool for companies registered in your industry in the last 90 days, and compare the list with what you get from the register yourself.
Questions
How fast does a B2B database go stale?
In our own Finnish data, 38.6 % of the records had closed after a year without rechecking. Contact details age faster than that, because people change jobs.
Why do stored databases miss new and small companies?
Because they are built from online presence, and new and small companies have little. The register has them from day one.
When is a stored database still better?
When you need large volumes of big, well-covered companies immediately. Many teams use both.