Methodology · Trust
Scope & Limits
Artmetria is built as a beta, designed as a method. This page explains what is available, what is partial, and how to interpret the platform responsibly.
Built as a beta, designed as a method
Artmetria is currently in beta. This means that some areas of the product are intentionally limited, progressively expanded, or still under review.
Rather than presenting incomplete coverage as certainty, Artmetria shows what is available, what is partial, and what should be interpreted with caution.
Our goal is not to produce a vague ranking of artists or artworks, but to build a structured analytical layer for looking, comparing and contextualising the art market.
Coverage is expanding, not universal
Artmetria does not claim to cover the entire art market.
At launch, coverage focuses on selected auction houses, categories and market segments where data can be monitored, normalised and interpreted with sufficient consistency.
Some sources may be watched without being fully indexed. Some houses may appear in the roadmap but remain outside the active scoring perimeter until the data quality is considered reliable enough.
When coverage is partial, Artmetria treats it as a methodological constraint, not as a hidden weakness.
Scores are analytical signals, not verdicts
The Artmetria Index and related scores are designed to organise market and object-level signals into a readable analytical framework.
They should not be understood as absolute judgements, price predictions, investment recommendations or guarantees of future performance.
A score may combine different layers of information: auction activity, liquidity, market visibility, object characteristics, confidence level, available provenance, literature, condition signals and other contextual factors.
When data is incomplete, the score should be read as provisional or limited. Artmetria is designed to make this distinction visible over time.
AI is used as an interpretive assistant, not as an authority
Artmetria uses AI to help structure, explain and contextualise information.
AI-generated outputs may assist with aesthetic profiling, artwork reading, artist discovery, market context and educational interpretation. They are not expert certificates, authentication opinions, legal opinions, tax advice or financial advice.
Where relevant, Artmetria aims to distinguish between descriptive information, interpretive commentary and market-related signals.
The AI layer is designed to support judgement, not replace it.
No investment advice
Artmetria does not provide financial advice, investment recommendations, buy/sell instructions or guarantees of value.
Any market signal, artist profile, index score, radar item or analytical note should be understood as educational and informational material.
Users should make independent decisions and, where appropriate, consult qualified professionals before making financial, legal, tax or acquisition-related decisions.
Data quality varies by source
Auction data may vary depending on the availability, structure and consistency of public information.
Some records may be incomplete, delayed, revised or unavailable. Sale totals, estimates, lot information, images, provenance, exhibition history and literature references may differ from one source to another.
Artmetria’s role is to normalise and contextualise information where possible, while preserving caution when the underlying source material is limited.
Some features are visible before they are fully available
Artmetria may display upcoming tools, Pro features or advanced analytical layers before they are fully open to all users.
This allows users to understand the direction of the platform while distinguishing between active features, beta features and roadmap items.
Features marked as beta, limited, coming next or Pro preview should not be interpreted as fully available production tools.
What improves over time
Artmetria is built as a progressive analytical system.
Its value increases as coverage expands, data quality improves, scoring layers mature, and more interpretive signals become available.
The beta is not presented as a finished authority, but as the first public version of a method.