How Market Insights works
The data, the method and the limits, in plain terms.
Market Insights turns the public domain name system into a picture of the hosting market. Every month we rebuild that picture from the ground up and freeze it as a numbered snapshot, so every ranking, share and trend on the site refers to the same point in time and can be compared fairly from month to month.
1 Building the map of domains
We start by assembling a very large list of registered domains across a wide range of top-level domains (the endings like .com, .net, or country domains such as .pl and .de). The aim is broad, representative coverage of the market rather than a claim to hold every domain that exists. This list is what everything else is measured against.
2 Seeing where domains really point
A domain's registration record is only a starting point. We query the live domain name system to see where each domain actually resolves today: which name servers answer for it, which address its website points to, and where its email is handled. When a domain stops responding, because it has expired or been parked, we keep its last-known details and simply mark it inactive, rather than pretending its provider suddenly lost it.
3 Understanding the servers
For the servers that domains point to, we collect general, publicly visible signals about the network they live on and the infrastructure that runs them. This is what lets us tell a large hosting network apart from a small independent one, and forms the raw material for attribution.
4 Attributing a domain to a provider
This is the heart of the method. Each domain is looked at from three independent angles:
DNS provider
Who runs the domain's name servers. This is our primary measure of a provider's size.
Hosting provider
Who owns the server the website actually runs on.
Email provider
Who handles the domain's incoming mail.
Each angle is decided by matching the technical fingerprints we observe against a large, human-curated set of rules that map those fingerprints to a known company. The most specific rule always wins, so a precise match beats a generic one. Anything we cannot confidently attribute is left as unassigned and excluded from the counts, rather than guessed. Because a domain's DNS, hosting and email can each be run by a different company, the three views often disagree, and that is expected.
Brand groups. Many providers operate under several brands. Where a parent company owns multiple operators, we roll them up into a single group, so a group's figures reflect the sum of all the brands it owns.
5 Detecting hosting control panels
Where possible, we identify the hosting control panel a server runs by looking for its characteristic signature from the outside. This is measured across the population of servers we can see, not across domains, and it only covers panels that are externally detectable. Large content-delivery networks are excluded, because they sit in front of many unrelated sites and would distort the result. Custom or panel-less setups are simply not counted.
6 Turning it into numbers
7 Who is winning and losing
By comparing snapshots over time, we can also see domains moving between providers: how many a provider gained, how many it lost, and who it gained them from or lost them to. Movers are filtered to a meaningful scale, so a tiny provider with an explosive percentage does not crowd out the real story.
What the numbers mean, and what they don't
- Size is measured by DNS. A provider's size here is the number of domains whose name servers it runs, which is not always the same as where the website is hosted or where the domain was bought.
- Country reflects the domain ending, not the owner's location, so markets dominated by global domains can look smaller than they really are.
- Some large numbers are technical. A few operators run huge volumes of automatically generated or parked domains; unusually large single-source figures should be read with care.
- It is a snapshot, not a registry. The picture is an automated estimate. It can differ from a provider's own internal figures, especially when a provider does not use its own name servers.
- History is still building. Recent months are the most reliable; long-range year-over-year comparisons will follow as more snapshots accumulate.
Market Insights applies the same automated method to every provider, with no manual ranking and no paid placement. The figures are published for information only.
Think your company is represented inaccurately?
We would genuinely like to fix it. The more we know about the domains and name servers you operate, the better we can attribute them.