What makes these numbers different
Every statistic below is a direct observation of the live web. We resolve domains, read the pages and mail records they publish, and count what is there, so each figure arrives with the population it was measured on and the date it was measured.
That is the opposite of a survey panel or a vendor questionnaire, and it means the numbers can be checked against the same public web you can reach yourself.
We publish no revenue market size and no growth forecast. Neither is observable from the web, and a number nobody can falsify is not evidence.
The population: how many domains, and which ones count
Three tiers, each defined by what we actually observed. The narrower the tier, the stronger the evidence behind every domain in it.
330M+ web-provisioned domains
Observed configured to serve the web at least once. A ceiling, not a count of current hosting customers.
185M+ web-active
Verified as web-serving on positive evidence only. What we cannot verify is disclosed rather than counted in.
116M+ confirmed-active business domains
The stricter floor: parked domains and dead pages removed. Headline claims use this tier.
Counts are published as floors, rounded down, so a claim stays true as the population grows. These are domain classifications, not counts of companies or legal entities.
Within the 13 European markets on this page, the confirmed-active business tier holds 18,542,389 domains. Of those, 15,440,157 resolve to a named hosting relationship — a provider we can identify as the one the customer buys from.
Consolidation: what the ownership record holds
We keep a curated ledger of ownership changes in hosting and domains, each entry carrying a source and a validation state. The counts below are the whole ledger in aggregate.
307 recorded events
Everything the ledger holds, including entries still being confirmed.
289 validated
Confirmed against a published source. Every figure in this section uses the validated set.
262 changes of owner
Acquisitions and mergers. The remainder are divestitures, rebrands and portfolio records, which are events but not purchases.
25 active buyers
Distinct acquiring groups across the validated set.
Of the validated deals, 70.9% bought core hosting, domains or the machines under them, and 29.1% bought software and adjacent services. That second share is the up-stack move: buyers adding what a hosting customer uses on top of the hosting.
250 validated deals carry an announcement date. Within the recent window the yearly flow runs 24 in 2019, 16 in 2020, 19 in 2021, 20 in 2022, 28 in 2023, 30 in 2024 and 28 in 2025, with 17 recorded so far in 2026 — a partial year, and a floor on it.
The ledger is a lower bound. A deal we have not sourced is not in it, and unannounced transactions never enter the public record at all.
This page stays at the aggregate. For which groups are buying and what happens to the books they buy, read how concentrated European hosting is and hosting's second act, or ask for the named detail through the connector.
The software these businesses run, and the base it is measured on
Across these 13 markets we have checked the web presence of 14,015,932 confirmed-active businesses directly — 75.6% of the 18,542,389 in the book. Not every one runs a website: a business can be reachable by mail and nothing more.
4,854,128 of them present a live company site or storefront. That is where the software a business runs is visible, so every share below is measured on those. The rest are generic pages, login screens, parked domains and sites that did not answer.
| Category | Business websites | Share of those measured |
|---|---|---|
| Content management system | 2,640,964 | 54.4% |
| Analytics or tag management | 2,013,434 | 41.5% |
| Ecommerce | 1,262,688 | 26.0% |
| Hosted website builder | 347,117 | 7.2% |
More than half of these business websites are built on a content management system, about two in five measure their own traffic, and about one in four sells online.
Applied to every business in the book, those percentages would overstate each of them.
If anything they understate what businesses run. We identify software by the traces it leaves on the page we read, so a product that leaves none counts as absent, and these shares cover every business site we could read rather than a sample of trafficked websites — which places them below the platform figures published from web-traffic panels.
Per-market technology figures, including the same categories against each market's own measured base, are on the market pages.
Email security posture across Europe
Measured on 18,684,696 European domains with mail exchangers observed. A domain that receives no mail has no posture to measure, so it is not in this base. Unlike the technology shares above, each figure here is a direct read of the records a domain publishes in DNS, where a policy is either present or it is not.
- 82.0% publish an SPF record, but only 29.5% use a hard fail. Another 47.3% use a soft fail, which asks receivers to accept the mail anyway.
- 49.7% publish a DMARC record. 9.5% publish one at the enforcing reject policy, and 31.4% stay at monitor-only.
Publication and enforcement are separate steps, and the gap between them is where the posture actually sits. The SPF without enforcement brief takes this apart market by market.
How this was measured
We resolve every domain in the population, read what its servers publish, and attribute the result to the operator behind it using ownership folds we curate by hand. Figures are refreshed weekly, and each one on this page comes from the snapshot dated at the top. Nothing here is modelled, projected or weighted to a panel.
The definitions behind each population, the coverage limits, the concentration readings and what we deliberately do not publish are all set out on the data and methodology page.
Common questions
Reproduce this — or ask it yourself
Every figure here is queryable through the HostingBrain connector. In Claude or any MCP-compatible assistant, this is the whole page in one prompt:
“Using HostingBrain, give me the European web hosting statistics I can cite: the domain population tiers with their definitions, the hosting-relationship HHI and effective owners for every market you measure, website technology adoption with the base it is measured on, European SPF and DMARC adoption, and the consolidation ledger in aggregate. State the snapshot date and the denominator for each.”
Resolves to market_concentration, technology_adoption,
email_security and consolidation_landscape (free tier).
Ask the follow-up yourself. HostingBrain answers questions like these with the date, the denominator and the limits attached, inside Claude and any MCP-compatible assistant.