In my opinion there are five primary elements which go into the value of a domain name:
- Brandability - Can the URL be turned into a brand?
- Type-In Trafic - How much traffic will come to the domain simply by virtue of the keyword(s) in the URL?
- Existing Traffic - How much traffic does the site generate, from what sources, and how stable are those sources?
- Monetization - What is the value (or probable value) of existing traffic?
- Age - When was the URL first registered? (Any expiration of the registration resets the clock.) How long has the website had content, indexed by the major search engines, in the desired subject area?
Not all factors are equally relevant, as is obvious from some huge money domain buys. SEO.com just sold for five million dollars, and porn.com reportedly fetched $9.5 million. They're simultaneously generic yet memorable, and are thus likely to generate both type-in traffic (people who type in "[keyword].com" into the browser's address bar), as well as return traffic by people who are pleased with the content due to the easily remembered domain name.
A brand need not have any automatic association with an industry to become extremely powerful. "Yahoo", "Google" and "Amazon" are powerful Internet brands, but the domain alone would tell you nothing about what you would find on their associated websites. If you can create your own brand and make it synonymous with your industry (or a significant subset of your industry) your brand should be extremely powerful and valuable. But most Internet ventures will not be blessed with that type of alignment of the stars. For "the rest of us", it is good to have a domain name that reflects the target industry and which is also brandable.
There are fewer domain names in this class than many people realize. Law.com is a great domain name, but I doubt that many people say to themselves, "I'm looking for a law, so I'll type in 'law.com' and see what I find." While a URL like "porn" says it all, there's no similar cohesion to the legal world - law.com, laws.com, lawyers.com, legal.com, attorney.com... there's no word which defines the industry. When you start using multi-word URL's you render the URL more and more generic with each added keyword. As a URL, Lawyers.com is worth vastly more than TrialLawyers.com, and Law.com is worth vastly more than my site, ExpertLaw. Both USLaw.com and USALaw.com seem like good URL's, but they are easily confused with each other. The former was a significant, venture capital-backed site which is now a pale shadow of its former self, the latter was a law firm sponsored site which now redirects to the law firm's webpage. In a testament to my larger point, USLaw.com incurred significant debt acquiring "LegalLink.com, a consumer site, AttorneyReferral.net and LegalPad.com" - does anybody remember any of those sites?
If you think you're sitting on a $million dollar domain name, be prepared to prove it. Buying a domain isn't much different in some senses from buying real estate. If you have a tumbledown shack of a website sitting on an established, brandable domain (an acre lot) on a valuable subject (one which can be monetized by the buyer) with a ton of natural traffic (hundreds of thousands to millions of unique visitors each month) from a variety of sources (a waterfront or oceanfront acre sized lot, a buyer will see a lot of value in the potential. If you have a domain name you think is great, but which has never been associated with a website, it's not so great. Find a way to at least get a tumbledown shack in place, or better yet a site with some real content and value. The best way to sell a website or URL is to be able to point to an actual earnings history. We all know what it means when we read about a property's potential in a real estate ad.