2026-09-30

How to appraise a domain before you buy it: a 20-minute worksheet

A five-line appraisal worksheet for domain buyers: label length, the radio test, category demand, TLD fit and history — plus the scaling that turns a score into a defensible offer, and the four checks that kill a deal.

Most appraisals fail at the first question. "What is this domain worth?" has no answer, because it hides the only thing that matters: who pays, how much they can pay without a meeting, and how long you can hold the name if nobody pays at all.

The names I regret buying were not overpriced against some abstract market. They were priced for a buyer who never existed.

Two numbers, not one

Every name carries a reseller number and an end-user number, and the two live in different rooms.

• Reseller number — what another investor pays knowing they must relist it, hold it, and clear a margin. In my own buying it lands at 20–35% of a plausible end-user price.
• End-user number — what a business pays because the alternative costs more: a rebrand, a keyword-free brand, or watching a competitor hold the exact term their customers type.

raleightreecare.com is a fair example from my own list. To a flipper it is a wholesale name. To the Raleigh tree-care company that would own it permanently, it is a lead source that closes jobs year after year. Same name, two honest prices.

The error is mixing them: using the end-user number to justify a wholesale-speed purchase, or using the reseller number to argue yourself out of a name with an obvious commercial buyer.

The five-line worksheet

I run the same five lines on every candidate. It takes about twenty minutes, and its only job is to stop the thing that ruins appraisals — liking the name first and hunting for reasons second.

1. Label length — characters before the dot. The median label in my portfolio is 11 characters and the shortest is 5. Every character past 12 is a discount you pay in recall and in typos.
2. The radio test — say it once on a call. Can the other person spell it back? If it needs repeating, you have just priced in your support cost for the next five years.
3. Category demand — does the label describe something people already search for? seattlestudioapartments.com has search intent behind it. fevoz.com has nothing behind it but your future marketing budget. Both sell; not to the same buyer, not at the same price.
4. TLD fit — a .com for lead generation and ecommerce; a .net or .co only when the product itself is the brand.
5. History — clean or not. This is the only line that can end the appraisal outright.

Score it before you price it

Score each line out of 20 and add them up. The total caps the price; it never sets it.

• 80–100 — Four figures is realistic to an end user with a business behind the name.
• 60–79 — Three figures to a reseller; four figures only with a named buyer in mind.
• 40–59 — Buy at wholesale only, and only if you enjoy holding inventory.
• Below 40 — Walk — unless you are collecting typos of somebody else's brand.

The brandability scorer in the toolkit applies this mechanically across length, pronounceability, radio test, spelling difficulty and memorability, and it shows the reason behind every point it deducts. Run candidates through it before you fall in love, not after.

Comparables, in this order

Comps are the part people fake. Look at them in order, and stop when the data stops agreeing:

1. Published sales of similar names — the strongest free signal there is.
2. Live listings with a buy-now price, when three or more cluster together.
3. Dead listings — a name priced high and unsold for two years is not a comp, it is a cautionary tale.

Then ask the seller the question that sorts professionals from storytellers: which sales are you comping this against? A professional answers in one sentence, with two or three names. Everyone else answers with the brand you could build. The counter-offer scripts for that conversation are in the blog archive.

The four checks that kill a deal

Each of these takes minutes, and each one has saved me money.

1. Redirect history. Check archive snapshots and the current redirect chain. A name that spent three years inside a link network still carries that reputation into your launch.
2. Typo leakage. If a dropped letter, a doubled letter or the plural lands on an active competitor, every email you send feeds them a little traffic. The typo and variant finder lists the realistic set, plus which of them are still unregistered.
3. Trademark. Search the label in the jurisdictions you actually sell in. A name you cannot use is a name you cannot sell either.
4. Renewal cost. Read the renewal price, not the promo price. Five years of a $40 renewal is $200 of hold cost that never shows up in an appraisal — and it is the line that turns a thin margin into a loss.

Turn the score into an offer

Here is the arithmetic, on one worked example. End-user band: $3,000, for a name that matches a service a mid-size company already sells. Reseller floor: 25% of that, so $750. Hold cost over five years: roughly $130 including the escrow fee on the eventual sale. If I expect eighteen months to find a buyer, my wholesale offer is $650–$800; my ceiling for an end-user purchase is a fraction of what a rebrand would cost that company.

The rule underneath it: never pay a price at which a single plausible buyer at your own buy-now number still leaves you under 30% gross. That one sentence filters more bad names than any scoring system I have used.

Walk away when

• The seller cannot name a single comparable sale.
• Two names in your own portfolio would do the same job for the same buyer.
• You would be embarrassed to say the label out loud on a call.

A twenty-minute appraisal, start to finish

1. Two minutes — length and the radio test.
2. Three minutes — search the head term and check whether the demand is commercial or vanity.
3. Five minutes — history: registration age, registrar moves, first archive snapshot.
4. Five minutes — typos and trademark.
5. Five minutes — comps, plus the renewal price, written down.

Then decide once: buy, counter with a number you can defend in one sentence, or walk.

Every name in my own portfolio — 154 of them, exact-match, brandable and geo — is priced with this worksheet before it gets a buy-now number, and every one transfers through escrow. If you want to see what those numbers look like on live inventory, the portfolio is where the whole list sits, no login and no "contact for pricing".

Originally published at https://dnwai.com/d/blog/appraise-a-domain-before-you-buy/ — 154 buy-now domain names, escrow transfers.