How do you value a domain name, and what makes one valuable?

You value a domain name by estimating it three ways and reading the gap between the results: what similar names have sold for, what income the name produces or could produce, and what it would cost a specific buyer to reach the same customers without it. What makes a name valuable is the combination of scarcity and usefulness: a short, clear, memorable name in a trusted extension, matching something people already look for or a brand someone wants to build, with a clean history and no trademark problem.

No single number comes out of that work. A domain name has a wholesale value, which is what another investor would pay for it today, and an end-user value, which is what a business that needs it would pay, and the two can sit far apart for the same name on the same day. An honest valuation gives a range, says which buyer each end of the range belongs to, and shows the assumptions behind it.

What makes a domain name valuable?

A domain name is valuable when it is scarce and when a business can turn it into customers, trust or savings. Scarcity alone is not enough: millions of names are unique and worth nothing beyond their registration fee, because nobody needs them.

No driver in the table below works alone. A short name in an extension nobody trusts, or a perfect keyword that infringes a trademark, fails on one line and loses most of the value the other lines gave it.

Two tests summarise most of that table. The radio test asks whether a person who hears the name once can type it correctly: no hyphen to explain, no digit that could be a word, no spelling to choose between. The pool test asks how many separate businesses could use the name without changing what they do. A name that passes both has many possible buyers, and the number of possible buyers is what turns a name's quality into a price.

How much do length and scarcity matter?

Length matters because the supply of short names is fixed and small, and the supply is what a buyer of a short name pays for. The limit comes from arithmetic, not from fashion.

The technical rules for names are set by the Internet Engineering Task Force. Its specification of the domain name system, known as RFC 1035, limits each label to 63 characters and describes the preferred syntax as letters, digits and hyphens. Within those rules the number of possible names at each length is a simple power, and it shows how quickly scarcity fades as a name gets longer.

The counts are per extension: 26 letters to the power of two, three and four, and 36 letters and digits to the power of two and three. They explain why very short names trade as a category of their own, valued partly as a scarce asset whatever the letters spell. From four letters upward the count is large enough that length stops carrying the value alone, and what the letters mean takes over.

For names made of words, length is measured in words more than in characters. A one-word name in a major extension is scarce because a language has a limited number of common commercial words. A two-word name is worth something when the two words form the natural phrase for a product or service. Three words and more are rarely worth more than registration unless they are the exact phrase a market uses.

How does the extension change the value?

The extension changes the value by changing how many buyers would accept the name, and the dot-com ending has the largest pool by a wide margin. The same word can be worth a large sum in one extension and the registration fee in another.

The size of each family of extensions gives a first measure of that pool. According to the Domain Name Industry Brief published by Verisign for the second quarter of 2025, there were 371.7 million domain name registrations across all top-level domains, of which 157.9 million were .com names, 143.4 million were country-code names and 39.5 million were names in the newer generic extensions. Dot-com alone was about 42.5% of every registered name.

Registration counts are not prices, but they show where businesses and their customers expect to find each other, and that expectation drives resale demand. Three practical consequences follow.

A country-code extension is valued inside its own country. A strong word in the extension of a large economy has a real market of local businesses; the same word in the extension of a small one has few buyers, unless the two letters of the extension happen to mean something else and attract buyers worldwide.

A newer generic extension has value mostly when the word and the extension read together as one phrase. Renewal cost matters more here than anywhere else: some registries price sought-after names at a premium renewal every year, and a name that costs hundreds of dollars a year to hold is worth less to any buyer than one that costs a few dollars.

A name in a secondary extension is partly priced against the dot-com version of the same word. If that version is unused and for sale, it sets a ceiling: few buyers pay more for the alternative than the original would cost them.

What do commercial intent, traffic and brandability add?

Commercial intent, traffic and brandability add value because each one is a route to revenue for the eventual buyer: a name that describes a paid product, a name people already type, or a name a company can build a brand on. They are also the three drivers that can be measured or at least tested.

Commercial intent is the difference between a word people are curious about and a word people buy under. The practical measure is what advertisers pay for a click on that term in search advertising: a high price per click means businesses already earn money from those words, and those businesses are the natural buyers of the name.

Traffic comes in two kinds. Type-in traffic, also called direct navigation, is the visitors who type the name straight into the address bar; it belongs to the name itself and passes to the next owner. Search demand is the number of people searching for the term each month; it does not come with the name, but it shows the size of the audience the name speaks to. Type-in visits are worth far more per visit than search volume, and they must be proven with the owner's own statistics, since a claim of traffic with no report behind it is worth nothing in a negotiation. Traffic left over from an old site fades, and traffic that exists because the name resembles a famous brand is a liability, not an asset.

Brandability is the hardest driver to measure and the one that produces the widest prices. A brandable name does not describe a product; it is short, distinctive and pleasant to say, so companies in many industries could adopt it. Its value depends on finding the one buyer who wants it, which is why two invented names of the same length can sell for very different amounts and why comparable sales are least reliable for this group.

How can a name's history or a trademark reduce its value?

A name's history or a trademark conflict can reduce its value to nothing, because either one can prevent the buyer from using what was bought. Both are checked before any price is discussed, and both belong in the valuation as a discount or as a reason not to value the name at all.

History covers what the name was used for before. Age by itself is a weak signal. What counts is whether earlier owners used the name for unsolicited mail, deceptive pages or other abuse, since blocklists and search engines may still treat it with suspicion; whether old links to it come from relevant sites or from link schemes; and whether the chain of ownership is clear, so that the seller can be shown to hold the name legitimately.

Trademark risk is the more serious of the two. Under the Uniform Domain Name Dispute Resolution Policy, adopted by ICANN and administered by providers such as the WIPO Arbitration and Mediation Center, a trademark owner can obtain the transfer or cancellation of a name by proving three things: that the name is identical or confusingly similar to its mark, that the registrant has no rights or legitimate interests in the name, and that the name was registered and is being used in bad faith. WIPO's guide to the policy states that a panel cannot award money, only transfer or cancellation, and lists among the examples of bad faith a registration made primarily to sell the name to the trademark owner for more than the registrant's out-of-pocket costs. The policy applies to generic extensions and to the country-code extensions that have adopted it.

National law can add money to the risk. In the United States, the Anticybersquatting Consumer Protection Act lets a court award statutory damages of not less than $1,000 and not more than $100,000 per domain name. Rules differ from one country to the next, and this is general information, not legal advice.

The risk is lowest for a common word used in its ordinary meaning, and highest for names that add a generic term to a known brand or misspell one.

What are the methods for valuing a domain name?

There are three methods for valuing a domain name: comparable sales, income, and replacement cost to the buyer. They are the three approaches that valuation practice applies to any asset, adapted to names.

The International Valuation Standards, published by the International Valuation Standards Council, name the three principal approaches as market, income and cost. The market approach compares the asset with similar assets for which price information exists. The income approach converts future cash flow into a single present value. The cost approach rests on what a buyer would pay to obtain an asset of equal usefulness. The same standards say that when approaches give divergent results, it may not be appropriate simply to average them, and that the valuer should look for the reason they differ.

Comparable sales work best where names are alike and trade often: short letter and number names, and descriptive names in a busy industry. Choose sales in the same extension, of the same structure, in a comparable industry, and recent enough to reflect current demand. Then adjust for the differences, as a property valuer adjusts for an extra room. Two cautions apply: many buyers require confidentiality, so reported sales are a partial sample, and a reported sale may be a wholesale trade between investors or a retail sale to a business, which are different markets.

The income method suits names that already receive visitors. The owner's net income from the name, usually advertising shown to type-in visitors less the renewal fee, is multiplied by the number of years of income a buyer is ready to pay for, or discounted year by year. It gives a floor, since a name that earns money is worth at least a multiple of what it earns, and it says little about a name with no traffic.

Replacement cost looks at the name from the buyer's side. A business has alternatives: a longer name, another extension, or its present name plus more advertising. The cost of the best alternative, including the visitors lost to a confusing address and the advertising needed to buy the same visits, is the most that business should rationally pay. This method gives a ceiling, and a different one for every buyer.

Why is the wholesale price so far below the end-user price?

The wholesale price is far below the end-user price because an investor who buys a name does not know when, or whether, the end user will appear, and pays for that uncertainty. The discount is the price of waiting, not a judgment on the name.

An end user buys a name to use it and values it on what it does for one business. An investor buys it to resell, and has to price in three things: the chance that a buyer comes within a reasonable holding period, the renewal fees paid every year in the meantime, and the commission of the marketplace or broker that handles the eventual sale. Multiply a retail price by a low annual chance of selling and the expected proceeds fall to a small fraction of that price. In the worked example below, the result is under a tenth of the retail value.

The practical consequence is that a valuation must say which market it describes. A seller who needs cash within weeks is selling at wholesale, to investors. A seller who can wait for a business that needs the name, alone or through a broker, is selling at retail.

Why do automated appraisals give different numbers?

Automated appraisals give different numbers because each tool is a statistical model trained on its own set of past sales, with its own choice of inputs, and none of them can see the buyer. Two tools can be reasonable and still disagree by a wide margin on the same name.

An automated appraisal typically reads the name's length, its extension, the words it contains, search and advertising data for those words, and the prices of names it judges similar. Several things limit what a model can produce from that.

• The training data differs from one tool to the next, and each set is incomplete, since most private sales are never reported.
• A model estimates a typical retail price. It does not estimate how long a sale would take, so it says nothing about wholesale value.
• A model cannot know that a funded company has just chosen this name for its brand, which is the event that produces the highest prices.
• Trademark conflicts and a poor history are often invisible to it, so a risky name can receive a confident figure.
• Brandable and invented names have few true comparables, and the estimates for them scatter the most.

Automated figures are useful for sorting a large portfolio in minutes and giving a first order of magnitude. They are a starting input to a valuation, never its conclusion. For a name of real value, a broker or an independent appraiser who trades in that category adds what the model lacks: knowledge of unreported sales and of who is buying.

How do you carry out a valuation step by step?

You carry out a valuation by clearing the legal and history checks first, then computing each method separately, then stating a range with the buyer each figure applies to. The example below does it for one name with every assumption stated.

The name is a two-word dot-com name, the natural phrase for a home service, held by an investor. Its statistics show 600 direct visits a month. Advertising shown to those visitors is clicked by 25% of them at $0.35 a click. Renewal costs $12 a year. All figures are illustrative.

Step four deserves a second look, because it is the calculation behind every wholesale offer. At a 1% annual chance of sale the same arithmetic gives about $346, and at 3% about $1,111. The name's quality hardly changes the retail figure in that test; what moves the wholesale value is how likely a buyer is to appear. For this name the income it already earns, $1,854, is higher than its resale expectation, so income sets the wholesale value.

How do you read the range a valuation gives?

You read the range by attaching each figure to the party it belongs to: the lowest figures are what an investor would pay, the middle one is what an average end user has paid for similar names, and the highest is the limit for one well-placed buyer. The figures answer different questions, which is why they are not averaged.

Three conclusions come out of the example. The wholesale value is about $1,854: an investor would not rationally pay much more, and an owner should not accept much less, since the name earns that back in three years. The retail asking price supported by evidence is about $9,750, roughly 5.3 times the wholesale figure. And for the regional company in step five, the name replaces $20,160 of advertising a year, about 2.1 times the comparable-sales figure, so that buyer could pay more than the comparables suggest and still come out ahead within a year.

A price agreed between that owner and that buyer would land somewhere between $1,854 and $20,160, and where it lands is a matter of negotiation, not valuation: who needs the deal more, what alternatives each side has, and how well each side knows the other's numbers. The valuation's job is to tell each party where its own limit is.

Two habits keep such a range honest. Deduct the cost of selling from any figure presented as proceeds: the 15% commission assumed here takes $1,462.50 from a $9,750 sale and leaves $8,287.50. And write the assumptions and the date next to the result, since click prices, traffic and demand all move.

When is a formal appraisal worth paying for?

A formal appraisal is worth paying for when a third party has to rely on the number: a court, a tax authority, an insurer, a lender, a co-owner or the other side of a company sale. For an ordinary purchase or sale between two parties, the range each party computes for itself is usually enough.

Typical cases are a dispute between co-owners, an estate, the contribution of a name to a company in exchange for shares, and a sale between related companies. In each, the reader of the report was not part of the deal and needs to see the method, the data and the reasoning.

A useful appraisal report states the purpose and the date of the valuation, the kind of value sought, the methods used and why, the comparable sales with their sources and adjustments, the traffic and income evidence with the reports attached, the legal and history checks carried out, and a range with a reasoned conclusion. The report should also say whether the appraiser is independent of the transaction.

Once a price is agreed, the transaction moves to its own questions: how the payment and the transfer of the name are sequenced so that neither party is exposed.