Germany's registry holds 18.0 million .de domains — DENIC publishes the counter. It is the number that opens every DACH hosting market-size slide. It is also, for anyone selling hosting, builders, email or SaaS, roughly twice the actual market: only 8.9 million .de domains have ever been provisioned to serve the web at all.
We calibrated our continuously observed web-provisioned population — every domain that has been set up to serve the web at least once, a deliberately generous ceiling that still includes parked pages and placeholders — against 15 official European registry totals. The result is consistent and large: the web-provisioned share of registered stock runs 49.5% to 78.7%.
The spread itself is the finding. Germany — Europe's largest registry — has the lowest web-provisioned share: its famous defensive-registration culture (register the name, the umlaut variants, the misspellings; host nothing) sits in the registry as stock no hosting product can address. The Nordics run the highest shares: fewer domains per business, more of them real.
The gap is not mystery inventory. It is registrar economics: defensive registrations, brand-protection portfolios, speculative names, and domains bought as an ancillary line on an invoice that never grew a website. That stock renews — it just never becomes a customer of anything beyond the registration itself. Which is why registrar books and hosting books are different assets even when the domain counts look identical — a distinction that matters rather a lot when books are being bought.
It also moves. Since our calibration five days earlier, Norway's share rose from 60.8% to 63.5% — not because Norwegians built 24,000 websites in a week, but because our harvest of the long tail landed. That is the honest mechanic of the number: a coverage measure improves as observation deepens; a registry total only counts renewals. We publish both dates so you can tell which is which.
Buyers of domain books. Hosting and registrar acquisitions price in an ARPU-expansion story: today the customer pays for a domain, tomorrow for hosting, a builder, mailboxes, security. But a never-web-provisioned domain cannot expand — there is no site to host, no mailbox to upgrade, structurally, and most of that stock never changes state. Two books of identical headline size can differ by thirty points in expandable share. If you paid a per-domain multiple built on the blended book, the never-provisioned half is a renewal annuity you priced as a growth asset.
Operators' own growth math. Every conversion benchmark with registered stock in the denominator is measuring against customers who were never in the market. A cross-sell campaign "reaching 100% of the book" wastes a quarter to a half of its spend by construction — and the wasted share is knowable, per book, in advance.
Vendors entering Europe. Country prioritisation inverts. Ranked by registry size, Germany is the obvious first market; ranked by web-provisioned reality, its 18M registry buys you 8.9M actual prospects — while the Nordics punch far above their registry weight. A registry-weighted go-to-market plan systematically overweights the markets where domains are registered defensively and hosted nowhere.
Lenders and diligence teams. When a target claims N domains, the first question is which N: registered stock, DNS-alive, web-provisioned, or confirmed-active business. These populations differ by construction — and revenue durability differs with them: defensive stock renews until a brand-protection budget gets cut; a hosted business site renews because a business depends on it. We label every count with which population it is.
Every figure here is queryable through the HostingBrain connector. In Claude or any MCP-compatible assistant, this is the whole brief in one prompt:
“Using HostingBrain, how complete is your view of European domains? Show the web-provisioned share of official registry totals per market — where is the gap between registered domains and the actual web biggest, and why?”
Resolves to coverage_calibration (free). See definitions('web_provisioned_universe').
Reproduce this analysis: the coverage_calibration
tool returns every benchmark above — our count, the registry total, the source URL, the date and the
confidence grade, per market, with history — free tier.
Ask the follow-up yourself. HostingBrain answers questions like this — with the date, denominator and caveats attached — inside Claude and any MCP-compatible assistant.