How to Value a Domain Name: A Practical Framework for 2026 Buyers and Sellers

Ask ten domain investors what a name is worth and you will get ten different numbers. Ask an automated appraisal tool and you will get an eleventh — usually wrong in a specific, predictable direction. The problem is that most public appraisal frameworks were built for the 2010s: keyword volume, exact-match SEO value, and TLD hierarchy. Those inputs still matter, but they are no longer the biggest levers on price.

This is the framework we use internally to price the names in our portfolio, and the one we recommend to founders and investors evaluating an inbound offer or a listed comp. It is designed to give you an anchor that stands up in a real negotiation — not a hopeful number.

Why automated appraisals are unreliable

GoDaddy's appraisal, Estibot, and their peers rely heavily on historical sales data plus keyword metrics. That produces two systematic errors:

1. They over-weight past comps for exact strings. A single 2019 sale of a similar name at $2,400 anchors every future estimate near $2,400, even if the market for that category has 5x'd. 2. They under-weight brand-fit and category legibility. An algorithm cannot tell that *Cortex.ai* is the obvious name for a neuroscience startup and *KrxTech.com* is not, even though the second has better raw keyword metrics.

Automated appraisals are useful as a floor for generic keyword names. For brandables — the bulk of the six-figure market — they are noise. Use the framework below instead.

The six attributes that actually predict price

Across ~800 tracked sales in our 2026 index, six attributes explain more than 80% of clearing-price variance:

1. Length (highest weight) Every character removed below 10 roughly doubles the value in the same tier. A 5-letter brandable .com is not 2x a 10-letter one — it is closer to 8–16x. This is the single most powerful variable in domain pricing and the reason 4L .coms clear five figures with no keyword content at all.

2. TLD (very high weight) .com is still the reference. .ai is the only extension that has genuinely closed the gap for AI-native categories — often trading at 60–80% of the matching .com in 2026, up from 30% in 2023. Everything else (.io, .co, .app, .xyz) trades at a steep discount for anything above the $2K tier.

3. Pronounceability (high weight) The one-read test: can an English speaker say the name correctly on first look with zero coaching? Names that pass this test add roughly 40% to comparable names that don't. Names that require any spelling ("is it -ly or -li?") lose that much and more.

4. Category legibility (high weight) Does the name map cleanly to a known vertical or product category? *Vertex* reads as enterprise infrastructure. *Neural* reads as AI. *Ledger* reads as fintech. Names with clear category legibility clear 40–55% higher than abstract coined names of the same length. Names that read as *two* categories at once (fintech AND AI, security AND enterprise) command the highest premiums.

5. Prior use and trademark residue (medium weight, usually negative) A name that previously ran a real, indexed product carries risk: existing trademark filings, DMCA history, or lingering search-result baggage. Investors typically discount 15–30% for meaningful prior use. A parked history is neutral.

6. Type-in and demand signals (medium weight) Verifiable inbound traffic, historical inquiries, and the volume of people typing the name directly into a browser. Most buyers ignore this and shouldn't. A name with even 20–50 monthly type-ins clears meaningfully higher than an equivalent name with none.

Attributes deliberately *not* on this list: exact-match SEO value, Google Trends score, TLD "prestige" beyond .com/.ai, and social handle availability. These matter, but they move price by single-digit percentages, not multiples.

A working formula

For a brandable .com or .ai in a recognizable category, we start from a category-median base price and apply multiplicative adjustments:

Estimated value = Base(category, tier) × Length multiplier × Pronounceability multiplier × Legibility multiplier × TLD multiplier × (1 − Prior-use discount)

The base price comes from tracked category comps (see our [2026 AI Domain Price Index](/blog/2026-ai-domain-price-index) for one worked example). Multipliers cluster in these ranges for a well-formed name:

• Length: 0.6x (11+ chars) → 1.0x (9–10) → 1.6x (7–8) → 2.4x (5–6) → 4.0x+ (≤4)
• Pronounceability: 0.85x (needs spelling) → 1.0x (readable) → 1.4x (one-read perfect)
• Legibility: 0.9x (abstract) → 1.0x (single category) → 1.5x (dual category)
• TLD: 1.0x (.com) → 0.75x (.ai) → 0.35x (.io) → 0.20x (.co) → 0.10x (rest)
• Prior use discount: 0 (parked) → 0.15 (light) → 0.30 (heavy trademark residue)

A 7-letter, one-read, dual-category .com with no prior use in the AI/ML infrastructure vertical: $8,400 base × 1.6 × 1.4 × 1.5 × 1.0 × (1 − 0) ≈ $28,200.

That is the number you should expect a serious buyer to clear — not the automated tool's $3,200.

Sanity-checking the number

Before you use the framework's output in a negotiation, run three checks:

1. Recent comps within the same tier. Look up NameBio for three sales of similar length, TLD, and legibility in the last 12 months. Your estimate should sit inside that spread, not above or below all three. 2. Marketplace BuyItNows. Search Afternic and Sedo for names one letter longer or one category adjacent. Working BuyItNow prices anchor the market's current willingness to pay. 3. The 24-hour test. Would you personally pay this number if you were building the business the name serves? If the answer is "no, but I'd pay half," your buyer will feel the same.

If any of the three checks disagrees with your framework output by more than 40%, trust the checks and revisit the multipliers.

Pricing a name you're selling

The framework produces a *fair market value* — roughly the median clearing price a motivated buyer will pay. When listing, most sellers price 20–40% above that number as an anchor, then negotiate down to it. Two exceptions:

• True 1-of-1 names (short, one-word, category-defining) list at 3–5x fair value and hold. The buyer pool is small, patient, and price-insensitive.
• Investment-grade names sold "price on request" signal that the seller expects a strategic buyer. This works when the framework value is above $50K; below that, price-on-request tends to depress inquiry volume.

Pricing a name you're buying

Two failure modes to avoid:

• Anchoring on the seller's ask. Sellers price above fair value on purpose. Ignore the ask; run the framework yourself; open at fair value minus 20%.
• Ignoring lease-to-own math. A $30K name over 24 months at ~$1,250/month is cheaper capital than the equity you'd give up naming it something worse. Most Afternic-listed premium names support lease-to-own — factor it into your budget conversation.

What good looks like

A negotiated purchase should land within ±15% of your framework's fair-value estimate. Landing 30% below suggests the seller was distressed or the name has a defect you missed. Landing 30% above suggests emotional pricing — either yours or theirs. Both should trigger a second look before wiring.

Where to go from here

If you're evaluating a specific name — inbound offer, listing you're considering, or a portfolio you're valuing — the two follow-ups worth reading are [How to Buy a Premium Domain: A Step-by-Step Buyer's Guide](/blog/buy-premium-domain-process) and the [2026 AI Domain Price Index](/blog/2026-ai-domain-price-index) for category-specific medians. For a walk-through on a specific name from our portfolio, [get in touch](/contact) — we'll show you the math.