It works as an easy-to-remember alias for complex numerical or alphanumeric IP addresses, letting you reach web servers through a browser without memorizing the underlying routing coordinates.
When you enter a name into a browser, the Domain Name System (DNS) resolves that string into a machine-readable address, and traffic is routed to the intended destination server. The domain name is the entry point into a global, distributed database that maps names to network locations.
That indirection is the whole point: services can move between physical servers or change their IP addresses entirely, and the address you type stays the same.
What is a domain name?
A domain name is a string of text that maps to an IP address, acting as the gateway for users to
reach websites through a browser. Computers identify each other using
Internet Protocol (IP) addresses — numeric dotted-decimal
IPv4 addresses such as 192.0.2.172, or longer hexadecimal IPv6 addresses. Those strings are efficient
for machines and impractical for people.
The usual analogy is a phone's contact list. A smartphone stores complex phone numbers under a searchable name, and DNS does the same job for the internet: you type human-friendly words into a browser, and the system resolves them into the numerical address needed to reach a particular web server.
Every website consists of two things — a domain name and a web server. The domain name is the street address that points a browser at the files, code, and content sitting on that server. Because the address is independent of the server behind it, you can replace the hardware without anyone noticing.
The parts of a domain name: reading right to left
Domain names are structured hierarchically and read from right to left, moving from general identifiers
to specific labels. The hierarchy actually begins at the root level, represented by an invisible dot
at the end of every domain — google.com. is the fully qualified form. Each part of the domain,
separated by a dot, describes its position within that hierarchy.
The section to the left of the root is the top-level domain (TLD), such as .com or .org. The label
immediately to the left of the TLD is the second-level domain (2LD), typically the unique name chosen by
the registrant. In google.com, .com is the TLD and google is the 2LD. More layered structures work
the same way: in google.co.uk, .uk is the TLD, .co is the 2LD indicating a commercial entity, and
google becomes the third-level domain (3LD). Similarly, developer.mozilla.org uses developer as a
subdomain to organize a distinct body of content.
Each label must follow formatting rules defined by core internet standards:
- Length: each label can be between 1 and 63 characters.
- Characters: only letters (A–Z), digits (0–9), and hyphens are permitted.
- Hyphens: a hyphen may not sit at the beginning or end of a label.
- Case: domain names are case-insensitive.
Types of top-level domain (TLD)
TLDs are categorized by intended use, geographic association, or eligibility requirements. Generic TLDs
(gTLDs) are the most common and include traditional extensions like .com for commercial use, plus
.net and .org. Many newer gTLDs have since been introduced for niche branding, among them
.app, .blog, .shop, .tech, .dev, and .io.
Country-code TLDs (ccTLDs) consist of two letters based on international country codes, indicating where
a site is registered or operates: .uk (United Kingdom), .jp (Japan), .au (Australia). Restricted
TLDs enforce stricter eligibility criteria — .gov is reserved for government departments, and .edu
is limited to educational and academic institutions.
Internationalized domain names (IDNs) allow non-Latin characters such as Arabic, Chinese, or Cyrillic
scripts. Because DNS is bound to the ASCII character set, IDNs use Punycode encoding to translate
non-ASCII characters into a standardized form beginning with xn--, which the existing global
infrastructure can process without modification.
Domain name, URL, website, hosting: what is the difference?
These four terms get used interchangeably, and they describe different things. A URL (uniform
resource locator) is the complete address used to find a specific resource online. In
https://cloudflare.com/learning/, https is the protocol, cloudflare.com is the domain name, and
/learning/ is the path to a specific page.
The rest is easiest to separate through a real estate analogy. Web hosting is the land — the server space where website files are stored. That publicly reachable computer supplies the resources that keep a site available. The website itself is the house: the collection of code, files, and images a visitor actually interacts with.
The domain name is the street address for that house. It is the identifier people use to find the location. A functional website needs both a place to host it and an address to reach it; neither one alone gets a visitor to your content.
Who manages domain names?
ICANN, the Internet Corporation for Assigned Names and Numbers, oversees the administrative hierarchy
and manages the global namespace. ICANN delegates management to domain registries — the
owners of TLDs such as .com — which in turn authorize registrars like GoDaddy or Cloudflare to lease
names to the public. Over 300 million domain names are registered today.
The system evolved out of the early internet, where developers manually mapped hostnames in a text file
called HOSTS.TXT. As the network grew, that approach stopped scaling. Paul Mockapetris invented DNS in
1983 to replace it, and responsibility initially sat with the Internet Assigned Numbers Authority (IANA)
before transitioning to ICANN in 1998.
Registration data lives in the WHOIS database, a public record you can query through a registrar's
search tool or the command-line whois utility. Querying a registered name like mozilla.org returns
its registrant — the Mozilla Foundation — along with creation date and domain status. Querying an
unregistered name such as afunkydomainname.org returns NOT FOUND, which is how you confirm a name is
free. Registrars require a valid real-world address at registration, and many offer WHOIS privacy
services to keep a registrant's personal contact details out of public view.
Registering a domain name: you are leasing, not buying
Registering a domain name is a lease, not a permanent purchase. You pay for the right to use a specific name for a period that typically runs 1 to 10 years. The current registrant holds priority for renewal, but the name returns to the market if the fees stop.
The expiration lifecycle includes a renewal window followed by a possible auction. A domain that is not renewed can also fall to domain squatting, where a third party registers the name in bad faith to resell it at an inflated price or to use it for phishing. To limit abuse, domains are subject to a 60-day transfer lock after initial registration, and moving a domain between registrars requires an authorization code from the current registrar.
Pricing runs from a few dollars a year to thousands, driven by market demand and by each registrar's business practices. Premium domains cost more because they are short or map closely to a specific industry. Choosing a trustworthy registrar matters more than the sticker price: some buy their own customers' expired domains the moment they lapse, then sell them back at a steep markup.
From a domain name to a website: how a DNS lookup works
A DNS lookup resolves a human-readable domain name into an IP address. When you type an address, the browser first checks its local cache. If nothing is found, the query goes to a recursive resolver, usually one provided by your ISP. That resolver then works down the hierarchy: root name server → TLD name server → authoritative name server.
The authoritative server holds the definitive answer and returns the final IP address. That answer travels back to the browser, which then connects to the host server. Caching and time-to-live (TTL) values keep this efficient; TTL sets how long a server may hold a record before it must be refreshed.
Because these records are distributed across servers worldwide, changes made at a registrar — such as pointing a domain at a new host — are subject to propagation delay. Servers have to wait for the old record's TTL to expire before they fetch updated data from the authoritative source, and that can take several hours. Nothing is broken during that window; the old records have not aged out yet.
How to choose a domain name
Picking a domain name is worth some thought, because it is often the first thing a customer judges you on. Keep it short and memorable, and avoid numbers, hyphens, and uncommon spellings that are easy to mistype. Relevant industry keywords help, as does a geographic marker if the business serves a specific area.
Beyond the primary address, registering additional extensions such as .net or .org protects a brand
from competitors and squatters. A domain of your own also gives you a matching email address —
info@yourbusiness.com rather than a free mailbox — and that detail carries real weight with customers
deciding whether a business is serious.
What should you learn next?
A domain name is only the addressing layer. To understand what runs underneath it — how packets travel, what routers and ISPs actually do, and why IP addresses exist in the first place — read how the internet works next.
From there, two natural directions are HTTP, the protocol a browser and server use to talk once the IP address is known, and hosting, where a website's files physically live.