NameBio has logged hundreds of thousands of domain sales. It’s the closest thing the industry has to objective comparable-sales data — but raw NameBio numbers, used naively, will mislead you. Here’s how to filter them.
What’s in the data
Every reported public sale: domain, price, date, marketplace. Going back to 2003. Filterable by length, TLD, characters, sale price band, year.
What’s NOT in the data: private off-market sales, broker-mediated transactions where the buyer requested confidentiality (most $100K+ deals), and sales reported later than the original transaction date.
The wholesale vs. retail split
This is the single most important filter to apply.
Wholesale sales happen on auction platforms (GoDaddy auctions, NameJet, SnapNames, DropCatch) and are mostly investor-to-investor. They reflect what the trade values a name at as a portfolio holding, not what an end-user would pay.
Retail sales happen on premium-listing marketplaces (Sedo, Afternic, brokered deals) and are mostly end-user purchases. The same name typically sells for 5-25× wholesale at retail.
When pricing a name to sell to an end-user, filter NameBio to retail sources only — otherwise you’ll anchor your asking price 10× too low.
The recency rule
A 4-letter .com sale from 2014 tells you nothing about 2026 prices. The .com market has shifted multiple times in the past decade. Filter to the last 24-36 months for any active pricing decision.
Older sales matter for pattern analysis (“is the 4L .com market trending up or down?”) but not for setting your ask.
The pattern match
You’re looking for comps that match your domain on at least two of: length, TLD, semantic category, and letter pattern.
A pure length+TLD match (any 5L .com) is a weak comp. A length+TLD+pattern match (any 5L .com starting with a vowel and ending in a consonant cluster) is stronger. A length+TLD+category match (any 5L .com in fintech branding) is strongest of all because it’s predicting the same buyer pool you’re selling into.
Outlier rejection
Drop the top 10% and bottom 10% of your comp set before computing a median. The top 10% are usually buyer-fit outliers (the company that needed exactly that name); the bottom 10% are usually distress sales or misreported amounts. The middle 80% reflects the actual market.
A concrete worked example
You own flux.io. You want to price it.
Step 1: Filter NameBio to .io sales, 4 characters, last 24 months. You get ~80 results.
Step 2: Filter further to retail sources (drop wholesale auctions). You’re down to ~25.
Step 3: Drop the top 3 and bottom 3 (outliers). 19 comps remain.
Step 4: Median of the 19. Let’s say it’s $4,200.
Step 5: That’s your retail floor. End-user asking should be 2-3× ($8K-$12K). Make-offer with min_offer_usd at $3,500-$4,000.
Step 6: If the buyer is clearly using the name as their primary brand (clear buyer fit), the upper range can stretch to $25-$50K. That’s the right-buyer multiplier — but you can’t price based on a buyer you don’t have yet.