Mean, Median, or Weighted? Choosing the Right Comp Average
Averaging your comparable sales sounds simple—until one number quietly distorts your entire valuation. Here's how to decide between mean, median, and weighted averages when pricing a domain.
You've pulled a clean set of comparable sales. The data looks solid, the matches are close, and now you just need one number to anchor your price. So you do the obvious thing—add up the sale prices, divide by the count, and call it your comp value. That single decision, made almost reflexively, is where a surprising number of valuations quietly go wrong.
The debate over median vs average domain comps isn't academic hair-splitting. The statistic you choose can swing your estimated value by thousands of dollars on the same underlying dataset. Choose poorly and you either leave money on the table or price yourself out of the market with a number you can't defend. Let's break down when each approach earns its keep.
What each average actually tells you
Before you pick a method, it helps to be precise about what each one measures. They are not interchangeable, and treating them as such is the root of most comp-averaging mistakes.
The mean (arithmetic average)
The mean is the sum of all sale prices divided by the number of sales. It uses every data point, which is both its strength and its weakness. Because it incorporates the full magnitude of every sale, a single unusually high or low price pulls the mean toward it. The mean answers: "What's the average dollar amount across all these transactions?"
The median
The median is the middle value when you sort every sale from lowest to highest. Half the sales fall above it, half below. It ignores the magnitude of outliers entirely and cares only about their position. The median answers: "What does a typical sale in this set look like?"
The weighted average
A weighted average lets you assign more influence to the comps that most resemble the domain you're pricing. A near-identical two-word .com sale might carry three times the weight of a loose, one-syllable-off match. The weighted average answers: "What's the average sale, adjusted for how relevant each one is to my specific asset?"
Why the mean betrays you more often than not
Domain sales data is notoriously skewed. Most transactions cluster in a modest range, and then a handful of trophy sales sit far out on the tail. That distribution is precisely the scenario where the arithmetic mean misleads.
Imagine five comps: $2,100, $2,400, $2,600, $3,000, and $41,000. The mean is $10,220. The median is $2,600. If you anchor to the mean, you'll walk into a negotiation with a number nearly four times what four of your five comps actually support. One whale sale hijacked the entire figure.
The mean assumes your data is symmetrical and outlier-free. Domain comps are almost never either.
This is exactly why the median is the sharper default for most domain valuations. It's resistant to the tail sales that plague comp sets, and it reflects the price a realistic buyer is likely to encounter. When someone challenges your number, "half of comparable sales came in above this and half below" is a far more defensible statement than an average that hinges on a single transaction. If you're wrestling with a distorting trophy sale, we've written a full playbook on handling outliers when one whale sale skews your comps.
When the mean is still the right call
The mean isn't useless—it's just situational. It earns its place when your comp set is genuinely tight and homogeneous. If you have eight sales of two-word tech .coms that all landed between $8,000 and $12,000, the mean and median will be nearly identical, and the mean captures the subtle spread the median discards.
Use the mean when:
- Your comps are closely clustered with no dramatic outliers.
- You have a healthy sample size—enough that no single sale dominates. Our guide on how many comparable sales you need for a defensible price is worth a read here.
- You want to reflect the full magnitude of every transaction, not just rank order.
In short: the mean rewards clean data and punishes messy data. Since domain data is usually messy, treat the mean as something you earn the right to use—not your starting point.
The weighted average: precision for serious valuations
For high-stakes appraisals, neither a raw mean nor a raw median fully captures reality, because not every comp deserves equal say. A weighted average solves this by letting relevance drive influence.
Consider weighting comps by factors such as:
- Similarity of the name — length, word count, syllable structure, and semantic closeness to your domain.
- Extension match — a .com comp should generally outweigh a .io comp when you're pricing a .com.
- Recency — a sale from six months ago reflects current demand far better than one from four years back.
- Sale channel and conditions — a fully brokered, arm's-length sale is more instructive than a distressed liquidation.
The tradeoff is transparency. A weighted average introduces subjective judgment, so you must document your weighting logic. A comp report that shows why each sale received the influence it did is enormously more persuasive than one that hides the math. If you're building a valuation you'll actually present to a buyer or acquirer, our walkthrough on how to build a comp report that justifies your asking price shows how to structure that reasoning.
A practical decision framework
Rather than picking a favorite statistic and forcing every dataset into it, let the data tell you which one fits:
- Start with the median as your baseline. It's the safest default for skewed domain data.
- Compare it to the mean. If the two are close, your data is well-behaved and the mean is fair to cite alongside the median.
- If the mean is dramatically higher than the median, you have outliers. Investigate them, decide whether they belong in the set, and lean on the median.
- For material acquisitions, build a weighted average on top of a vetted comp set—and show your work.
Whatever you choose, the quality of your underlying comps matters more than the averaging method. A perfect weighted average of bad comps is still worthless. Sourcing matters: our overview of where to find reliable comparable sales data covers vetted sources like NameBio, and it's worth pairing with our rundown of the five comparable-sales mistakes that wreck valuations. If you're staring at a name with no clean comps at all, we have a fallback playbook for pricing without comparable sales.
The strategic takeaway
Choosing between mean, median, and weighted average isn't about statistical purity—it's about defending a number under pressure. The median protects you from the tail sales that distort domain data. The mean rewards you when your comps are genuinely tight. And a documented weighted average gives you the precision serious acquisitions demand. Master all three, and you stop being at the mercy of whichever number happened to come out first.
If you're evaluating a specific name and want to pressure-test where it should sit against real comparable sales, browse the curated inventory at PixelWorks Domains—or reach out about a particular acquisition. We're happy to talk through the valuation logic behind any asset before you commit capital to it.
