Methodology

DOMYRA's valuation model is deliberately conservative. It publishes a price range only when real comparable sales support one, and always explains what it used.

1. Normalisation

Input is trimmed, lower-cased, stripped of protocol and paths, converted to its ASCII (punycode) form for internationalised names, and checked against DNS label rules. Syntactic validity says nothing about whether a domain is registered or available — DOMYRA does not check availability.

2. Naming quality score (0–100)

A deterministic score from length, hyphens, digits, dictionary-word structure, pronounceability, commercial keywords and extension. It starts at 50 and each factor adds or subtracts points shown in the result. It is not a price.

3. Comparable sales

Only verified (weight 1) and reported (weight 0.6) sales in USD count. Asking prices and demo data are always excluded. Each sale is weighted by similarity (extension, length, shared words, structure) and recency (half-life 3 years; ignored after 8 years). Sales below 35% similarity are dropped.

4. Estimate

With at least 3 comparables, the central value is the weighted median of comparable prices, adjusted by naming quality (clamped between 0.5× and 1.8×). The range uses weighted quartiles, widened as confidence falls. With fewer comparables, the result is Insufficient data.

5. Confidence

Confidence grows with the effective weight of comparables, is capped at 85%, and is reduced when evidence is reported rather than verified. Labels: Low (<30%), Moderate (30–60%), High (≥60%).

What the model does not yet use

Search volume, industry demand, traffic, trademark risk and currency conversion. These require authorised data sources that are not connected. Every result lists these limitations.

Current limitations

No sales-data provider is connected yet, so there are currently no comparable sales and estimates show Insufficient data. DOMYRA does not check registration or availability, does not give legal or trademark advice, and does not predict resale prices. Opportunity scores and Market Pulse use the same evidence rules and stay empty without licensed data.

Versioning

Every evaluation stores its inputs, factors, comparables and model version so past results can be audited after the model changes.