Reading comparable domain sales properly

A comparable sale is useful when it answers a narrow question: what did a similar name, in the same extension, sell for, to a similar buyer, at a similar time. Every one of those four conditions is routinely ignored, and each failure moves an estimate in the same direction, upwards.

Matching the term type

The first requirement is that the compared names belong to the same category. An exact commercial keyword, a short letter combination, a brandable and a numeric string are separate markets with separate buyers. A sale in one says nothing about pricing in another, no matter how similar the character counts look.

Within a category, the match has to go further. For keyword names, the sector matters: a term in a trade where customers are expensive to acquire supports prices that the same length of term in a low value sector does not. For brandables, syllable count and construction style matter more than meaning. Getting this wrong is the most common error in amateur appraisal, and it is the error automated tools make as well, as discussed in valuation tools and limits.

Matching the extension

Prices do not translate between extensions, and no fixed ratio exists between them. A sale in the main commercial extension is not evidence for a national country code equivalent, and a country code sale in one country is not evidence for another country's equivalent.

The relationship also depends on the state of the stronger name, which is why cross extension comparisons need the context described in the extension effect page.

Matching the date

Aftermarket prices move with the funding environment for new companies, with advertising budgets, and with the availability of credit to investors. A sale from a period of loose funding is not evidence for a tighter one, and vice versa.

Recency should be weighted, but not to the point of using a single recent transaction. One sale is an observation about one buyer.

Matching the buyer

The largest source of spread in observed prices is who bought. Investor to investor trades sit at wholesale levels. End user purchases sit far above them. Brand protection purchases sit somewhere else again, driven by legal risk rather than by the name's usefulness.

Public records rarely identify the buyer type, and this is the single biggest limitation of the visible data. Where a marketplace records a sale without context, the same figure could represent either end of that range. Some inference is possible: a name resold quickly after purchase suggests a wholesale trade, and a name that immediately begins operating as a live site suggests an end user.

What the public record leaves out

The visible sales lists are a sample, and the sampling is not random.

• Private and brokered deals are frequently subject to confidentiality terms and never appear. These skew heavily towards the largest transactions.
• Reporting is voluntary and selective. Venues publish what suits them. Weak results are less likely to be volunteered than strong ones.
• Low value bulk sales are captured comprehensively at some venues and not at all at others, which distorts any average calculated across sources.
• Failed and expired listings are invisible. Names that sat unsold for a decade leave no record, so the visible data contains outcomes and not attempts.
• Auction results reflect the bidder pool present on one day, which for expiring names is mostly investors rather than end users.

The combined effect is a record that overstates both the typical price and the probability of a sale. Where the records come from and what each source includes is set out in sales comps sources.

A list of sales is a list of the transactions that happened and were disclosed. It is not a list of what names are worth.

Using comps without being misled

Work with a set rather than a single sale, and prefer the middle of that set to its average, since a small number of very large transactions will drag any mean upwards. State the reasoning behind a chosen figure in terms a buyer could check: term type, extension, period, and why the compared names are genuinely similar.

Then treat the result as a range with an explicit floor rather than a target. In a negotiation, the comparable evidence a seller can produce is far more persuasive than an assertion of worth, but only if it holds up when the buyer examines it. A poorly matched comp presented as proof damages credibility for the rest of the conversation, which is why the preparation described in handling inbound offers starts with evidence rather than with a number.