Why Domain Pricing Looks Random
You see one four-letter .com sell for $200,000 and another for $1,200. You see a category keyword sell for $50,000 and a longer phrase sell for $800. The randomness is real β but it follows rules. Once you understand the seven factors that drive domain value, the pricing stops feeling arbitrary.
This is the same framework professional brokers, investors, and automated appraisal tools use. The difference is they apply it implicitly; here we make it explicit so you can use it yourself.
The Seven Factors That Drive Domain Value
1. Length
Shorter is more valuable, almost always. The progression is exponential, not linear:
- 1β2 character .com: $1M+ (essentially all sold at this point)
- 3-character .com: $50Kβ$500K depending on letter combination
- 4-character .com (LLLL): $1Kβ$50K
- 4-character .com (LLNN, NLLN, etc.): $200β$2K
- 5-character .com (LLLLL): $300β$5K
- 6+ character word/phrase .com: wide range, driven by other factors
2. Memorability and Pronounceability
"Zillow" is more valuable than "Zfqxlw" even at the same length. Domains people can pronounce on first read, repeat without spelling, and type from memory command premiums. This is the "radio test" applied to valuation.
Approximate weight: a pronounceable 5-letter .com may be worth 5β10Γ a random 5-letter .com.
3. Keyword Value
Generic keyword domains β "Insurance.com", "Hotels.com", "Loans.com" β sell for millions because they describe an entire market category and bring inherent search and direct-navigation traffic. The hierarchy:
- One-word category .com (Insurance.com): $1Mβ$50M
- Two-word category .com (CarInsurance.com): $100Kβ$5M
- Niche keyword .com (PetInsurance.com): $20Kβ$200K
- Long-tail keyword .com (CheapPetInsurance.com): $1Kβ$20K
4. TLD
The same name is worth dramatically different amounts at different TLDs. As a rule of thumb:
- .com baseline: 100%
- .io / .ai (in tech): 30β60% of the .com value, sometimes more for AI-relevant names
- .net / .org: 10β25% of the .com value
- .co: 15β30% of the .com value
- New gTLDs (.app, .xyz, .shop, etc.): typically 1β10% of the .com value, though category-fit names buck this
5. Brandability
A "brandable" name is one that a startup could imagine using as a company name β distinctive, ownable, trademark-friendly. Brandability is partly subjective, but markers include:
- Two-syllable cadence (Spotify, Stripe, Notion)
- Distinctive consonants (often "X", "Z", "Q")
- Made-up words rather than dictionary words
- Easy trademark path
Strongly brandable 5β7 letter .com names regularly sell for $5Kβ$50K.
6. Comparable Sales
The most reliable single signal is what similar domains have actually sold for. Public sales databases (NameBio is the standard) record reported sales going back decades. Look up sales of:
- Same length and TLD
- Same syllable count
- Same general category
- Same time period (last 24 months β older sales reflect different market conditions)
Three to five recent comparable sales are more useful than any algorithm. They tell you what the market actually pays.
7. Existing Use and Backlinks
If a domain has been actively used, it may have:
- Existing backlinks from other websites (potentially valuable for SEO if the link profile is clean)
- A traffic history (cached search rankings, direct-navigation users)
- Existing trademark associations (potentially valuable, potentially a liability)
Backlink and traffic data adds real value but requires careful evaluation β a "high-DA" domain with a spammy link profile is worth less than a clean domain at the same metrics.
How Automated Appraisal Tools Work
Tools like EstiBot, GoDaddy Appraisal, and Saw.com run a regression model trained on past sales data. They weigh the factors above and output a single number. Three things to know:
- They are accurate within a band, not at a point. A $5,000 appraisal might mean "real value somewhere between $2,000 and $12,000". Use it as a sanity check, not a price tag.
- They underperform on outliers. Truly premium one-word .com domains and truly worthless typo domains both fall outside the training distribution. Models tend to over-correct toward the mean.
- They cannot read context. A domain that perfectly matches a recent IPO's company name has value the model cannot see.
The Buyer's Pricing Strategy
If you are buying:
- Run two or three automated appraisals to get a baseline range.
- Look up 5+ comparable sales on NameBio. Disregard the appraisal if real comps disagree.
- Set a maximum price before you reach out to the seller.
- Make a reasonable opening offer (typically 30β50% of your max). Sellers expect negotiation; opening at your max gives you nowhere to move.
- Walk away if they will not move into your range. There is almost always another viable name.
The Seller's Pricing Strategy
If you are selling:
- List on a marketplace (Sedo, Afternic, Dan, etc.) at 1.5β2Γ the appraisal mid-point. Most domains do not sell at the asking price; the asking price is an opening anchor.
- Be patient. Liquid premium .com domains may sell in months; specific niche domains sometimes take years to find their buyer.
- If a buyer reaches out unsolicited, value goes up. Inbound interest is a signal you have a name someone specifically wants.
- Decide in advance whether you accept payment plans (Lease-to-own structures common on Dan.com) β they widen your buyer pool but tie up the asset.
The Honest Bottom Line
Most domains are worth less than their owners think and more than their buyers want to pay. The gap is normal. Automated appraisals plus recent comparable sales close most of that gap. Negotiation closes the rest.
If you ever feel certain about a domain's value, look up a few real sales β you will probably calibrate down. If you ever feel a domain is worthless, look up a sale from the same category β you may calibrate up. The data is public; the discipline is using it.