Methodology
Last updated 03 Jun 2026
Every Artlex composite score combines six independent signals, each measuring a different facet of an artist's momentum, validation, and risk. The composite is a synthesis, not a prediction. The exact source weighting and model formula are proprietary, but every published score includes a rationale and a coverage breakdown so you can judge each signal on its merits.
What follows is a plain-English description of each category, what it captures, and where it falls short.
The six signals
Cultural velocity
Measures recent multilingual cultural footprint: how often an artist is referenced across art-world conversation, in what languages, and whether the reference trend is rising, stable, or fading. A high cultural-velocity score means the artist is in active circulation; a low one means they aren't currently visible, not that they aren't culturally significant.
Press signal
Measures the volume, recency, and authority of editorial coverage. Tier-1 international art press carries more weight than aggregator blogs; recent coverage carries more weight than archived. Press signal alone is not enough. It is the noisiest of the inputs and is deliberately balanced by the others.
Institutional signal
Measures museum acquisitions, public-collection holdings, biennial invitations, and curatorial validation. Slow-moving but high-quality. An artist with strong institutional signal has typically been validated by people whose job is to be careful.
Market signal
Measures secondary-market activity where it exists: verified auction lots, sale-vs-estimate performance, lot frequency, price recency, and gallery-tier representation on the primary market. For emerging or primary-market-only artists, this signal is intentionally thin or absent rather than fabricated.
Medium trend
Measures whether the medium an artist works in is currently gaining or losing collector and curator attention. A strong artist working in a fading medium is read differently from the same artist riding a wave.
Risk-adjusted signal
Penalises thin data, single-source concentration, and short-term hype spikes that aren't supported by institutional or editorial anchors. This is a data-fragility signal, not an authenticity assessment. Artlex does not detect forgery, misattribution, or fake provenance. A high composite score paired with a strong risk penalty reads very differently from the same composite without one, and both are surfaced.
Confidence and data depth
Every score carries a confidence indicator derived from how many signals fired and how much underlying evidence supports them. A score built on multiple converging signals with substantial press, exhibition, and institutional evidence is labelled high-coverage. A score built on a handful of press mentions with no institutional or market anchor is labelled thin and flagged as preliminary.
Treat thin-data scores as direction, not destination. The confidence label is there to keep that distinction visible.
What Artlex does not measure
The score is about an artist, not a specific artwork. Artlex does not assess, verify, or guarantee:
- Authenticity of a specific work
- Provenance or chain of ownership
- Legal title or seller authority
- Physical condition or conservation history
- Edition integrity for prints or multiples
- Fair market value of an individual work
- Tax, import/export, or resale-restriction implications
- Counterparty AML or sanctions exposure
For any of the above, commission a qualified specialist before transacting. Artlex is a research input, not a substitute for due diligence.
Known coverage gaps
Coverage is uneven, and pretending otherwise would mislead. Today:
- Western and English-language sources are over-represented; multilingual coverage is expanding but still partial in regions like Latin America, MENA, and parts of Asia and Africa.
- Secondary-market coverage varies by auction house and region. Some major international houses are fully indexed today; others, plus most regional houses, are partial or pending. Market signal on an artist whose record sits primarily with a not-yet-integrated source will read lighter than the artist's actual liquidity warrants.
- Private-sale activity is by definition invisible to a public-data system and is not represented.
- Artists with little public digital footprint (no indexed press, no public exhibition records, no encyclopedic reference coverage) score thinly even when culturally significant. The score reflects what is measurable, not what is true.
Updates and corrections
Scores re-compute on a regular cadence as new evidence arrives. Material methodology changes (adjustments to how signals are weighted, or new signals being added) are noted on this page when they ship.
Factual errors (wrong gallery, wrong school, mistaken identity, outdated exhibition record) can be reported via the factual-error form. We read every report and correct what we can. Reports about facts are separate from disagreements with a rating. We fix factual errors, but we don't negotiate scores.
In short:Artlex scores an artist's momentum, validation, and risk from a blend of cited public, licensed, and curated data. The exact formula is private; the reasoning behind every score is public. Use it as one input among several, not as a replacement for due diligence.