What is my domain worth? A method you can apply yourself

A domain is worth what one specific buyer will pay on one specific day. No book value, no ticker, no clearing price. That doesn't make valuation a guessing game, though. You can build a range you'd happily defend in an afternoon, using four inputs: comparable sales, a short list of adjustment factors, the buyer's business case, and a liquidity discount that reflects how long you're actually willing to sit on the thing.

Start by accepting the unglamorous baseline. Most registered domains are worth the renewal fee and not a penny more. Too long, wrong extension, a hyphen, a digit, an invented word nobody types into anything — any one of those can pin a name at wholesale forever. If you work through the method below and conclude "this is a £200 name", the method worked. You've just saved yourself two years of waiting for £20,000 that was never on its way.

Comps are evidence, but only if you read them like a sceptic

Public databases such as NameBio record reported transactions with price, date and venue. Marketplace listings tell you what sellers hope for, which is a much weaker signal and should be treated as such.

Hunt for comps along three axes:

• Same keyword. Holding something with "ledger" in it? Find every reported sale containing "ledger". That tells you whether the word itself carries commercial weight or just vibes.
• Same pattern. Two-word .com. Dictionary word .com. Four-letter pronounceable. Industry plus "ly". Patterns price far more consistently than individual words do.
• Same buyer type. A name only a funded startup would touch behaves nothing like a name every plumber in the country might want.

Then interrogate what you've collected. How old is it? A 2015 sale is history, not evidence; weight the last 24 to 36 months heavily. Where did it sell? An expiry auction figure is a wholesale price paid by a reseller who fully intends to mark it up, whereas a brokered end-user sale is retail, and the gap between the two runs 5–20x for the same class of name. Averaging them together produces a number that describes nothing. And was it actually cash? Reported totals sometimes represent a payment plan stretched over years, which is worth less than the same figure today.

Gather eight to fifteen genuine comps, bin the top and bottom outliers, and look at what's left. Your starting band will be wide. Something like £3,000 to £25,000 is entirely normal at this stage. Narrowing it is what the rest of the process is for.

The six things that genuinely move the number

Work through these in order. Each nudges you up or down inside the comp band.

Extension. The .com remains the default and prices like it. A strong .co, .io or .ai can command real money in the right sector, but for the identical string most alternative extensions trade at a meaningful discount to .com, often a small fraction of it. Country extensions price against their local market rather than the global one: a .co.uk is valued by British buyers, full stop.

Length and syllables. Shorter is better, but say-ability beats character count. Seven letters someone can read down the phone without spelling it out will outperform a five-letter jumble.

Commercial intent. "Insurance", "loans", "clinic", "software" — these attach to industries where a single customer is worth a great deal. "Sunset", "willow" and "pebble" are pretty and generic. A word's value tracks the profit per customer in the industry it serves.

Friction. Every hyphen, digit, doubled letter and ambiguous spelling takes a bite. Hyphens are the most brutal, frequently a 70–90% discount against the clean equivalent.

Brandable or exact-match. Two entirely different buyers. Exact-match keyword names sell on search and category authority; invented brandables sell on availability, trademark headroom and how well they stretch. Work out which one you own and comp accordingly, because the buyer pools barely touch.

Encumbrances. If your string maps onto somebody else's registered trademark, your buyer pool shrinks to that one company, and they may fancy a legal route rather than a cheque. Practical value: close to zero.

Now ask what it does for the person writing the cheque

Comps tell you what the market has paid. The business case tells you how your price will feel to a real buyer, and therefore where in the range you should sit.

• Name three likely buyers. If you can't, your pool is thin and your price should say so.
• What do they turn over? A figure that's rounding error for a Series B company is impossible for a two-person agency. Businesses tend to treat a domain as brand capital expenditure sized against a year of marketing spend, not against a hosting bill.
• What's the alternative? If a decent variant is sitting there for £12, you're not holding much of a hand. If every sensible option in the category is gone and yours is the last clean one standing, the number goes up.
• What does walking away cost them? A buyer already trading under the name, or one with printed packaging in a warehouse, has considerably more to lose than a curious founder.

This is why the same string sells for £4,000 to a freelancer and £40,000 to a funded company. You aren't pricing letters. You're pricing fit.

Three numbers, and only one of them is public

Retail ask. The top of your adjusted band. It's what you list at, and it assumes you'll wait for the right buyer, which may mean years. Most premium names attract a handful of serious enquiries annually, so patience isn't part of the strategy. It is the strategy.

Realistic settlement. Usually 60–80% of the ask once someone has negotiated. If that figure still pleases you, your ask is sound. If it doesn't, your ask is too low.

Walk-away floor. Wholesale: what another investor would hand you today, commonly 10–25% of retail. Know it before you open a conversation and never let a buyer find it.

Then apply a liquidity discount, because money now beats a hypothetical larger sum later. If three more years of renewals and attention is the price of waiting, and the retail buyer may simply never turn up, an offer at half your ask today can be the better decision. That's a judgement about your circumstances, not about the name.

What the algorithms are and aren't for

Automated valuations parse length, extension, keyword frequency and past sales. They're genuinely useful for two jobs: triaging a large portfolio quickly, and telling you which names you've been kidding yourself about. What they can't see is the only thing that sets price, namely whether a real buyer with real budget wants this specific name this year.

Treat the output as one comp among many. Inside your hand-built range, mild confirmation. Wildly outside it, go back and check what you missed, then trust your own work.

Price is what you agree; proceeds are what you keep

Two adjustments to build in before you commit to a floor.

Commission. On Names.com, sellers pay 15% on a completed sale and nothing if it doesn't complete; buyers pay no fee. Set your floor on the net, not the headline.

Transfer risk. Buyers discount for uncertainty, and they should. When the transaction runs through escrow, with neither side paying or transferring first, that discount largely evaporates because nobody is pricing in the chance of losing their money. A marketplace that handles this is worth more than the commission it charges.

One last discipline. Write the valuation down alongside the comps that produced it. When an offer lands at a third of your ask, you'll want the reasoning on the page in front of you rather than somewhere in your memory, competing with the offer.