Web Hosting Statistics Updated November 2026

Human Written & Fact Checked · Updated

100 Domain Statistics for 2026

A domain name is an address, not proof of an active website. Read registration statistics with that distinction and the geography becomes much more useful.

This collection combines global registration evidence with detailed German and French registry geography and one Canadian registrar's naming study. The measurements retain their own populations and periods instead of being blended into a hosting-market estimate.

Headline figures

  • Global domain registrations: Total domain registrations: 411.6 million (Q3 2026). (DNIB / Verisign, Oct 2026)
  • Global domain registrations: .com registrations: 169.8 million (Q3 2026). (DNIB / Verisign, Oct 2026)
  • Global domain registrations: Preliminary .com/.net renewal percentage: 74.3% (Q3 2026). (DNIB / Verisign, Oct 2026)
  • North Rhine-Westphalia, Germany: North Rhine-Westphalia: 3,363,094 .de domains (Year-end 2025). (DENIC, Jul 2026)
  • Hamburg, Germany: Hamburg: 328.6 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)
  • Saxony-Anhalt, Germany: Saxony-Anhalt: 87.8 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)
  • Île-de-France, France: Île-de-France: 27% of .fr domain stock (rounded) (2025). (Afnic, Mar 2026)
  • Corse, France: Corse: 6.0% annual .fr stock change (2025). (Afnic, Mar 2026)
  • .fr registry: Total .fr stock: 4,319,120 registrations (2025). (Afnic, Mar 2026)
  • .fr registry: DNSSEC-signed .fr domains: 21.6% (2025). (Afnic, Mar 2026)
  • WHC registrations, Canada: .ca represented approximately 73% of WHC's .ca/.com sample (2020–2025). (Web Hosting Canada, Jun 2026)
  • WHC registrations, Canada: Letters-only names: 90% (2020–2025). (Web Hosting Canada, Jun 2026)

How large is the registered-domain market?

Domain stock is not a website count. Names can be parked, redirected, unused or attached to several services, so registration totals cannot stand in for hosting accounts.

How quickly did global registrations change?

  • Total registrations grew 2.5% quarter over quarter (Q3 2026). (DNIB / Verisign, Oct 2026)
  • Total registrations grew 8.7% year over year (Q3 2026). (DNIB / Verisign, Oct 2026)
  • New-generic-TLD registrations: 56.6 million (Q3 2026). (DNIB / Verisign, Oct 2026)

The overall stock and the newer generic-extension category describe different levels of the same system. A registration can persist without an actively maintained website. Quarterly and annual changes also use different baselines, so their rates should not be added. The figures establish the scale of registered naming infrastructure; they do not identify how many people registered names, how much traffic those names receive or how much hosting revenue they generate. Domain portfolios can contain defensive registrations and multiple addresses pointing to the same destination.

What happened to established generic extensions?

  • Combined .com/.net registrations: 182.5 million (Q3 2026). (DNIB / Verisign, Oct 2026)
  • Combined .com/.net quarterly growth: 1.9% (Q3 2026). (DNIB / Verisign, Oct 2026)
  • Combined .com/.net annual growth: 6.1% (Q3 2026). (DNIB / Verisign, Oct 2026)
  • .net registrations: 12.6 million (Q3 2026). (DNIB / Verisign, Oct 2026)
  • New .com/.net registrations: 13.1 million (Q3 2026). (DNIB / Verisign, Oct 2026)
  • Prior-year-quarter new .com/.net registrations: 10.6 million (Q3 2025). (DNIB / Verisign, Oct 2026)

The combined stock includes both long-established generic extensions. New registrations are a flow during the quarter, whereas the registered base is a stock at quarter end. Renewal estimates describe yet another process and retain the publisher's preliminary label where applicable. A higher new-registration count does not establish higher retention or more active websites. The prior-year comparator below concerns new registrations in the corresponding quarter, instead of the previous quarter. Keep these distinct accounting windows visible when citing a change.

What does the country-code category measure?

  • Country-code registrations: 151.4 million (Q3 2026). (DNIB / Verisign, Oct 2026)
  • Country-code quarterly growth: 1.9% (Q3 2026). (DNIB / Verisign, Oct 2026)
  • Country-code annual growth: 4.6% (Q3 2026). (DNIB / Verisign, Oct 2026)

Country-code extensions form a naming category, not a reliable map of server locations. Registration policies and usage patterns vary by registry; a name under an extension need not imply that its content is hosted in that country. The category aggregates stocks reported through the brief's stated source system. Its growth rate cannot be interpreted as the growth in local businesses, internet users or national hosting capacity. For geographic analysis, holder-location records such as the German and French registry series below are more direct, although they still do not identify server location.

Where are German .de domain holders located?

Registry geography describes the holder's recorded location. State totals reveal the distribution of registered names without claiming that a domain's server or audience sits in the same state.

How many .de names are assigned to southern and western states?

  • Baden-Württemberg: 2,081,445 .de domains (Year-end 2025). (DENIC, Jul 2026)
  • Bavaria: 2,699,117 .de domains (Year-end 2025). (DENIC, Jul 2026)
  • Hesse: 1,300,817 .de domains (Year-end 2025). (DENIC, Jul 2026)
  • Rhineland-Palatinate: 665,802 .de domains (Year-end 2025). (DENIC, Jul 2026)
  • Saarland: 139,645 .de domains (Year-end 2025). (DENIC, Jul 2026)

These states are presented as comparable rows from the same registry snapshot. A holder can control several names, and an organisation's registered address can concentrate a portfolio in one location. The counts therefore measure domain holdings instead of distinct businesses or residents. They do not measure website visits, hosting capacity or the number of operating online shops. Keep the state labels and reporting year attached when quoting an individual value, especially when comparing these counts with a population-normalised measure.

How do the northern states and city states compare?

  • Berlin: 929,612 .de domains (Year-end 2025). (DENIC, Jul 2026)
  • Bremen: 134,775 .de domains (Year-end 2025). (DENIC, Jul 2026)
  • Hamburg: 627,674 .de domains (Year-end 2025). (DENIC, Jul 2026)
  • Lower Saxony: 1,525,598 .de domains (Year-end 2025). (DENIC, Jul 2026)
  • Schleswig-Holstein: 561,455 .de domains (Year-end 2025). (DENIC, Jul 2026)

A city state is an administrative unit in this table, just like a larger territorial state. Its smaller geographic area does not change the definition of a registered name. Comparing absolute counts is useful for locating portfolios, but it does not control for population or economic structure. The same person or company can register multiple names, and a portfolio's administrative address may differ from its operating locations. Consequently these values should not be described as the share of residents with websites or as the local web-hosting market.

What do eastern territorial states report?

  • Brandenburg: 317,774 .de domains (Year-end 2025). (DENIC, Jul 2026)
  • Mecklenburg-Western Pomerania: 177,607 .de domains (Year-end 2025). (DENIC, Jul 2026)
  • Saxony: 548,177 .de domains (Year-end 2025). (DENIC, Jul 2026)
  • Saxony-Anhalt: 191,430 .de domains (Year-end 2025). (DENIC, Jul 2026)
  • Thuringia: 232,374 .de domains (Year-end 2025). (DENIC, Jul 2026)

The table applies the same registration-location method across these states. Smaller totals cannot by themselves establish weaker digital participation: population, company location and portfolio ownership all affect the count. These records also cover only the national extension in this series. Names held under other extensions are outside this comparison. A regional digital-economy claim would need evidence about usage, businesses and residents beyond the registry. Here the observation remains a dated stock of registered names attached to holders in the named state.

How does German domain density vary by population?

Population-normalised domain holdings answer a different question from total portfolio size. The numerator remains registered names, so density is not the percentage of residents who own a domain.

What are the densities of city states and northern states?

  • Berlin: 245.8 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)
  • Bremen: 194.9 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)
  • Lower Saxony: 186.9 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)
  • Schleswig-Holstein: 189.3 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)
  • Brandenburg: 123.1 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)

The population denominator comes from the provider identified in DENIC's table. Dividing names by residents makes differently sized states easier to compare, but it does not turn portfolio holdings into individual adoption. Companies may hold large portfolios, and residents can hold more than one name. Administrative addresses also matter. The accompanying registry map uses the same metric at district level, exposing local variation that a state average can conceal. Its colour bands should not be read as an additional state ranking.

How do southern and western states compare on density?

  • Baden-Württemberg: 183.6 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)
  • Bavaria: 200.9 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)
  • Hesse: 202.6 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)
  • North Rhine-Westphalia: 184.9 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)
  • Rhineland-Palatinate: 159.5 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)
  • Saarland: 140.4 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)

The unit is names per population denominator, not a household penetration rate. Differences may reflect company headquarters, portfolio concentration or other features that the registration table does not separately measure. No causal explanation is assigned here. A comparison with total registrations can show how population adjustment changes a region's position, but does not establish which region has better infrastructure. The series contains no direct measurement of connection quality, hosting performance, sales generated by websites or residents' digital skills.

What densities appear in the eastern territorial states?

  • Mecklenburg-Western Pomerania: 109.0 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)
  • Saxony: 134.1 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)
  • Thuringia: 109.5 .de domains per 1,000 residents (Year-end 2025). (DENIC, Jul 2026)

Use these observations as descriptive registry statistics. They concern a particular extension and its holders' recorded locations, not all names that residents or firms may use. A low value does not mean that an equivalent share of people lack internet access or websites. The source population series and the domain snapshot must remain paired if a later edition is compared with this one. Changes in either numerator or denominator can alter density. The separate takeaway identifies the remaining state in this regional comparison without repeating its figure here.

How is France's .fr domain stock distributed regionally?

A large regional stock and rapid regional growth are different signals. Afnic reports both, allowing the geography of existing holdings to be separated from the direction of change.

Which regional shares describe the northern and eastern mainland?

  • Grand-Est: 5% of .fr domain stock (rounded) (2025). (Afnic, Mar 2026)
  • Auvergne-Rhône-Alpes: 12% of .fr domain stock (rounded) (2025). (Afnic, Mar 2026)
  • Bourgogne-Franche-Comté: 3% of .fr domain stock (rounded) (2025). (Afnic, Mar 2026)
  • Centre-Val de Loire: 2% of .fr domain stock (rounded) (2025). (Afnic, Mar 2026)
  • Hauts-de-France: 5% of .fr domain stock (rounded) (2025). (Afnic, Mar 2026)
  • Normandie: 3% of .fr domain stock (rounded) (2025). (Afnic, Mar 2026)

These percentages use total extension stock as the denominator, instead of a subtotal limited to metropolitan France. Foreign holders and overseas territories therefore matter to the full distribution. Source shares are rounded to whole percentages, which limits precise ranking of regions with similar displayed values. A registration is linked to a holder location; it does not locate the hosting server or establish where visitors live. The table should be used to describe the registry's geography, with no inference about regional internet penetration.

What shares do western and southern regions hold?

  • Nouvelle-Aquitaine: 7% of .fr domain stock (rounded) (2025). (Afnic, Mar 2026)
  • Bretagne: 4% of .fr domain stock (rounded) (2025). (Afnic, Mar 2026)
  • Corse: 0% of .fr domain stock (rounded) (2025). (Afnic, Mar 2026)
  • Occitanie: 8% of .fr domain stock (rounded) (2025). (Afnic, Mar 2026)
  • Pays de la Loire: 5% of .fr domain stock (rounded) (2025). (Afnic, Mar 2026)
  • Provence-Alpes-Côte d'Azur: 8% of .fr domain stock (rounded) (2025). (Afnic, Mar 2026)

The smallest displayed share rounds to zero; it does not mean that the region has no registrations. Rounded regional shares should not be converted back into exact counts by multiplying them by national stock. The original report supplies counts separately when that precision is required. These rows describe the location of holders under this extension, not all domain names used by French organisations. They also do not measure the share of regional businesses with a website. Portfolio concentration can shape the distribution without an equivalent change in website activity.

Which regions grew or contracted during the year?

  • Grand-Est: 2.0% annual .fr stock change (2025). (Afnic, Mar 2026)
  • Nouvelle-Aquitaine: 3.5% annual .fr stock change (2025). (Afnic, Mar 2026)
  • Auvergne-Rhône-Alpes: 2.5% annual .fr stock change (2025). (Afnic, Mar 2026)
  • Bourgogne-Franche-Comté: 1.9% annual .fr stock change (2025). (Afnic, Mar 2026)
  • Bretagne: 2.9% annual .fr stock change (2025). (Afnic, Mar 2026)
  • Centre-Val de Loire: 2.6% annual .fr stock change (2025). (Afnic, Mar 2026)
  • Île-de-France: -0.2% annual .fr stock change (2025). (Afnic, Mar 2026)
  • Occitanie: 3.1% annual .fr stock change (2025). (Afnic, Mar 2026)
  • Hauts-de-France: 3.2% annual .fr stock change (2025). (Afnic, Mar 2026)
  • Normandie: 5.2% annual .fr stock change (2025). (Afnic, Mar 2026)
  • Pays de la Loire: 0.2% annual .fr stock change (2025). (Afnic, Mar 2026)
  • Provence-Alpes-Côte d'Azur: 2.9% annual .fr stock change (2025). (Afnic, Mar 2026)

Stock changes use each region's own earlier stock as the comparison base. They are not percentage-point changes in regional market share. Registration movement does not establish business formation or hosting revenue.

What does .fr registry operation data reveal?

Registry operation metrics describe naming infrastructure instead of website success. Retention, signing and registration term each require their own definition.

How do stock growth and maintenance differ?

  • Annual .fr stock growth: 2.4% (2025). (Afnic, Mar 2026)
  • .fr maintenance rate: 82.2% (2025). (Afnic, Mar 2026)
  • Foreign holders: 11.9% of .fr stock (2025). (Afnic, Mar 2026)

The maintenance rate is the registry's measure of stock retained over the period; Afnic explains that it is not interchangeable with a renewal rate. New registrations and deletions affect total stock growth, while maintenance focuses on the persistence of the existing base under the report's method. Foreign-holder share is a separate location measure and does not locate servers. None of these metrics tells us whether a registered name carries an active website. Keep the source's exact definitions when comparing another extension, because registry operational measures are not necessarily harmonised.

How common are internationalised and multi-year registrations?

  • Internationalised .fr domains: 0.8% (2025). (Afnic, Mar 2026)
  • Multi-year .fr registrations: 10.7% of stock (2025). (Afnic, Mar 2026)

Internationalised names and longer registration terms describe different attributes and can overlap. Their percentages should not be added as if they partitioned the registry. An internationalised name concerns the naming format, not a reliable measure of the language used on a website. A multi-year term describes the registration commitment, instead of a hosting contract or a guarantee that a site remains active. The section's signing chart concerns authentication of DNS data; it does not measure the encryption of website traffic, incident frequency or the security of hosted applications. These measures concern the registry layer and require separate evidence about any website's operation.

What naming patterns appear in a Canadian registrar's data?

A registrar cohort can illuminate actual naming choices without representing the entire country. WHC's evidence is strongest when its customer base and observation window remain explicit.

Which extensions were registered within WHC's cohort?

The multi-year totals and the annual rows cover different windows. Do not add them together or describe the annual figures as the whole country's registrations. The publisher excludes the incomplete current year from the percentage calculations, keeping the completed-year cohort comparable. These are registrations processed at this registrar, not all names held by Canadians. Transfers, existing portfolios elsewhere and names under other registrars are outside the stated measurement. Extension preference in this cohort cannot establish a national market share or prove that one extension performs better in search.

How long and complex were sampled domain names?

Length excludes the extension. The average and the share below a threshold describe different properties of the same distribution. A name containing a digit can also contain a hyphen, so these categories need not be mutually exclusive. The observed naming pattern does not test memorability, conversion or search ranking. Those outcomes would require a separate experiment or usage study. These figures are descriptive evidence from the registrar's completed-year registration cohort, with its customer mix intact.

What do language and spelling classifications indicate?

  • Approximately 45% of sampled .ca names read as English (2020–2025). (Web Hosting Canada, Jun 2026)
  • Approximately 11% of sampled .ca names read as French (2020–2025). (Web Hosting Canada, Jun 2026)
  • Approximately 5% of sampled .ca names read as bilingual (2020–2025). (Web Hosting Canada, Jun 2026)
  • WHC found 221 names containing centre (Study dataset since 2020; report June 2026). (Web Hosting Canada, Jun 2026)
  • WHC found 54 names containing center (Study dataset since 2020; report June 2026). (Web Hosting Canada, Jun 2026)

WHC infers language from short strings and explicitly labels the classification directional. A coined brand may have no dictionary classification, and a bilingual-looking name does not establish the registrant's language or the language of the eventual website. Spelling occurrences count strings in the reported study, not people or nationally representative preferences. Their exact full-study window is less explicit than the completed-year percentage denominator, so the inventory preserves that qualification. No inference about the performance of English, French or bilingual websites follows from these naming observations.

Glossary

Methodology and sources

Every figure on this page comes from the organisation that produced it: regulators, statistics offices, public bodies, research institutes, and companies reporting their own data or surveys. Each one links to the original report or dataset and carries its release date. Survey results keep their population and fieldwork period, and financial figures keep the measure the producer used. Figures from different producers are not added together or ranked against each other. The page is reviewed monthly and updated when a producer releases newer data.