As domain names became interesting to marketers because of their advertising and marketing potential, rather than just being used to label Internet resources in a technical fashion, they began to be used in manners that in many cases did not reflect the intended purpose of the label of their top-level domain. As originally planned, the structure of domain names followed a hierarchy in which the TLD indicated the type of organization (commercial, governmental, etc.), and addresses would be nested down to third, fourth, or further levels to express complex structures, where, for instance, branches, departments and subsidiaries of a parent organization would have addresses in subdomains of the parent domain. Also, hostnames were originally intended to correspond to actual physical machines on the network, generally with only one name per machine.

As the World Wide Web became popular, site operators frequently wished to have memorable addresses, regardless of whether they fit properly into the structure; thus, because the .comdomain was the most popular and therefore most prestigious, even noncommercial sites began to obtain domains directly within that gTLD, and many sites desired second-level domain names in .com, even if they were already part of a larger entity where a subdomain would have been logical (e.g., abcnews.com instead of news.abc.com).

Shorter, and therefore more memorable, domain names are thought to have more appeal. As a convenience, methods were implemented to reduce the amount of typing required when entering a web site address into the location field of a web browser. A website found at ''http://www.example.org'' will often be advertised without the http://, since the HTTP protocol is implicitly assumed when referring to web sites. In many cases, web sites can be also be reached by omitting the www prefix, as in this given example. This feature is usually implemented in DNS by the website administrator. In the case of a .com, the website can sometimes be reached by just entering example (depending on browser versions and configuration settings, which vary in how they interpret incomplete addresses).

The popularity of domain names also led to uses which were regarded as abusive by established companies with trademark rights; this has become known as cybersquatting, in which a person registers a domain name that resembles a trademark in order to profit from visitors looking for that address. To combat this, various laws and policies were enacted to allow abusive registrations to be forcibly transferred, but these were sometimes themselves abused by overzealous companies committing reverse domain hijacking against domain users who had legitimate grounds to hold their names. Such legitimate uses could include the use of generic words that are contained within a trademark, but used in a particular context within the trademark, or their use in the context of fan or protest sites with free speech rights of their own.

Laws that specifically address domain name conflicts include the Anticybersquatting Consumer Protection Act in the United States and the Trademarks Act of 1999 in India. Alternatively, domain registrants are bound by contract under the UDRP to comply with mandatory arbitration proceedings should someone challenge their ownership of a domain name.

Often email phishing scams will abuse subdomain names to appear to be a legitimate site. For instance, an email might purport to be from Bank of America, and include a link to a fake login screen hosted on http://www.bankofamerica.com.abc.def.ghi.jkl In this case, the actual domain is ghi.jkl, but appears at first glance to be bankofamerica.com.