Installed infrastructure reflects the past. New websites reveal momentum.

HostingBrain analyst brief · July 2026 · cohort 2026-04-22 → 2026-06-17 · anonymized, aggregate-level

Every hosting investor asks some version of the same question: are hyperscalers and edge platforms capturing the new websites? Installed market shares can't answer it — they average over a decade of old decisions. So we measured arrivals instead: every European domain that entered our web-active population over eight complete weeks — ~69,000 new web-active domains across 17 European markets — compared against the 18M-domain installed base, by hosting-front class.

The first cut looked dramatic. The second cut is the story.

The trap: momentum that wasn't

Raw, the European numbers scream edge-migration: 26% of new domains sit behind a CDN/edge front, against 12% of the installed base. Doubling! In one market — Italy — 74% of new domains arrived CDN-fronted.

Except: behind Italy's surge sits a single aftermarket operator that moved roughly 10,000 domains behind a CDN in one window — 70% of the entire Italian new-domain cohort, one nameserver. Not ten thousand Italian businesses choosing the edge; one drop-catch portfolio changing configuration.

CDN/edge-fronted share, European web-active domains Installed base 12.2% New — raw the mirage 26.2% New — bulk events screened 14.1% screening = markets where one nameserver holds >30% of the new cohort are excluded (one market flagged: a ~10k-domain aftermarket move behind a CDN in a single window)

Single operators move portfolios. Markets move slowly. If you don't screen bulk events, you will mistake one for the other.

The organic picture: arrivals look like the installed base

With bulk events screened out (~55,000 organic new domains), the distribution of new European websites across hosting-front classes is almost indistinguishable from the installed base:

share of European web-active domains (bulk events screened) new (8 weeks) installed base Traditional hosting origin 77.8% 78.4% CDN/edge-fronted 14.1% 12.2% Hyperscaler origin (visible) 8.1% 9.3% ~55k organic new domains vs 16.9M installed · 17 European markets · snapshot 2026-07-05 shares of domains with a classifiable front; 'hyperscaler' = visible AWS/Google/Azure origin

Traditional hosting holds ~78% of new European websites — the same share it holds of the installed base. The CDN/edge front is mildly ahead among arrivals (+2pp). And visible hyperscaler origins mildly under-index on new sites (8.1% vs 9.3%).

Why this contradicts the vibe

Three mechanics inflate naive "the cloud is taking the new sites" reads:

1. Placeholder pollution. Fresh domains disproportionately sit on registrar forwarding and placeholder infrastructure — which runs on major-cloud IPs. To an ASN-level read, a parked new domain looks like a cloud-hosted new website. Our web-active tier excludes placeholder and forwarding infrastructure by reverse-DNS fingerprint, so that inflation never enters these numbers.

2. Bulk events. One operator, one window, ten thousand domains — screened here, headline-making if not.

3. The masking bound, stated. For ~14% of organic new sites the origin sits behind a CDN and is not visible. What serves behind the edge is a bound, not a guess — if every masked origin were a hyperscaler, the picture would change; history suggests it is a mix. We report what we can see, and how much we can't.

What this means

For operators: the feared leak of new customers to hyperscalers is not visible in European arrival data. The front door of new websites still overwhelmingly opens on traditional hosting — the new-customer fight is still winnable at acquisition, which is precisely where acquisition strategy shapes the monetisable base.

For investors and advisors: momentum claims — scary or rosy — need placeholder-clean, event-screened arrival data. A single-window, single-market read can show you a 74% "edge surge" that is one portfolio changing nameservers. Screen first, conclude second.

Method & caveats. New = first observation of a domain in HostingBrain's continuously observed web-provisioned population during the trailing eight complete weeks (discovery-paced — when we first see a domain serving, not its legal registration date). Basis: web-active domains (parking and forwarding placeholders excluded by reverse-DNS operator fingerprint). Classes: CDN/edge-fronted (origin masked), visible hyperscaler origin (AWS/Google/Azure), traditional (other visible origin); unresolved origins excluded from shares. Bulk-event screen: markets where one nameserver holds >30% of the new cohort are flagged and excluded from the organic aggregate — the flag itself ships in the data. Global aggregates are sensitive to which TLD cohorts arrive in a window; per-market and screened-Europe scopes are the reliable reads. Aggregate and anonymized; no operators named.

Reproduce this analysis: the hosting_momentum tool returns every number above — global, Europe raw vs screened, and per-market with bulk-event flags — and it is available on the 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.

Get free access