DATA QUALITY & MARKET BENCHMARKS · NFT TEAMS

Turn raw NFT transactions into usable market intelligence.

Atelier ZH cleans and structures marketplace transaction data, calculates collection-level price benchmarks, and delivers recurring reports your team can review, explain, and reuse. Built for teams that need consistent market data without building and maintaining their own analytics pipeline.

View a sample report Discuss your data scope

Currently available for OpenSea transactions on Ethereum.

WHAT THE PIPELINE RECEIVES, EVERY DAY
ETHWETHMANA price = 0missing valuesSANDDAI
01
Retrieve
02
Clean
03
Calculate
04
Deliver
Collection benchmark — example collection 1.42 ETH
The problem

Marketplace data is available. Usable market data is harder to obtain.

NFT marketplaces generate large amounts of transaction data, but raw records are rarely ready for analysis. The same market may contain several currencies, duplicated events, zero-value transactions, missing information, unusual token scales, and activity that shouldn't be treated as a standard sale.

Currencies
Dozens of currencies, no common unit
ETH, WETH, MANA, SAND, DAI and more: without normalization, transactions aren't directly comparable.
Data quality
Records that don't reflect a real sale
Transfers disguised as sales, mis-scaled tokens, zero-value transactions, duplicates: raw records need review before they can be analyzed.
Teams that work directly with this data face the same questions every time: which transactions to include, how to normalize prices, which unusual values are still valid, and whether the same method can be reproduced next month. Atelier ZH turns these decisions into a documented, repeatable process.
What we deliver

One managed workflow, three usable outputs

Not a menu of features to pick from — a single process, with three things you actually receive at the end of it.

01
Clean transaction data
We retrieve and process marketplace transactions using fixed, documented rules. Currencies are normalized, duplicates are handled, problematic records are flagged, and every transformation stays traceable.
  • Clean CSV or database export
  • Daily or weekly updates
  • Documented inclusion / exclusion rules
  • Processing and anomaly logs
02
Collection-level benchmarks
We calculate historical market references at collection level, using only information available at the time. Each benchmark comes with context: transaction count, price dispersion, activity, and available history.
  • Median price and price range
  • Volume and liquidity indicators
  • Reliability indicator per estimate
A consistent analytical reference — not a guaranteed resale price for an individual NFT.
03
Recurring monitoring & reports
We turn the processed data into reports and dashboards your team can use directly — trading volume, price distribution, concentration, and unusual movements worth a closer look.
  • Weekly report or live dashboard
  • Charts ready for internal or client use
  • Unusual movements highlighted for review
  • Plain-language note on what changed
Current coverage

A defined scope, processed consistently

OpenSea Ethereum

Atelier ZH currently supports historical and recurring transaction analysis for OpenSea on Ethereum. Additional marketplaces or chains can be assessed as part of a custom project — availability depends on data access, transaction structure, required history, and update frequency. We only confirm a new source after its data quality and processing requirements have been reviewed.

Illustrative example

What a collection benchmark looks like

A worked example, not live data: a collection's real transaction prices against the calculated benchmark. Hover over the chart to explore a point.

Real transaction price Collection benchmark
Illustrative collection · 24-day window Price in ETH
Example result

What a before / after looks like

An anonymized run from one of our datasets, illustrating what changes between the raw feed and the delivered output.

Input
  • 3,842 raw transaction records
  • 3 currencies (ETH, WETH, USDC)
  • 127 zero-value records
  • 46 duplicate events
  • 31 unusual price observations
After processing
  • 3,638 transactions retained
  • All prices normalized to ETH
  • Excluded / flagged records documented
  • Daily benchmark recalculated
Output
  • Weekly market report
  • Clean dataset export
  • Price & volume dashboard
  • Reliability indicator
Interpretation — Trading volume increased month over month, but the apparent rise in average price was concentrated in a small number of unusually high transactions. The median benchmark remained comparatively stable.
Method

A consistent number requires a consistent process

Comparable over time
The same cleaning and calculation rules are applied at every update, so reports from different periods stay comparable.
Traceable decisions
Excluded, corrected, and flagged records are logged. Your team can review how a result was produced.
Historically valid
Benchmarks only use information available at that date. Future transactions are never used to improve past results.
Reliability shown with the result
A benchmark from a handful of transactions isn't presented the same way as one from a deep, active market. History and dispersion are shown alongside every reference.
Who it's for

Built for teams that use market data but don't want to maintain the pipeline

Research & reporting teams
Use consistent figures, documented methodology, and ready-to-publish charts in recurring market studies.
Digital-asset projects
Track activity, price distribution, liquidity, and concentration without maintaining a separate analytics workflow.
Advisory & professional services
Support client reports with processed transaction data and a methodology that can be reviewed and explained.
Custom data teams
Add a cleaned, structured NFT transaction layer to an existing internal workflow through scheduled exports or a dedicated integration.
Pricing

Pricing based on data scope, not a feature count

Scope is defined together before work begins. The amounts below are indicative starting points for a standard scope.

Pilot analysis
From €350
To evaluate the methodology on one collection and a defined historical period.
  • One collection or defined dataset
  • Data-quality review
  • Cleaned export + benchmark example
  • Short findings report and review call
Discuss a pilot
Most common
Managed monitoring
€900 / month
For teams that need recurring, consistent market information without maintaining their own pipeline.
  • Defined collection scope
  • Daily data processing
  • Weekly monitoring report
  • Reliability & anomaly indicators
  • Email support
Discuss your scope
Custom integration
On request
For teams that need the processed data inside an existing product or system.
  • Custom asset & source scope
  • Scheduled export or dedicated API
  • Additional marketplace feasibility study
  • Integration & maintenance support
Contact us

Final pricing depends on the number of collections, historical depth, update frequency, data volume, and delivery format.

Contact

Tell us the scope, we'll tell you what's possible

zhengzhihong191@gmail.com

Share which collection(s), what time range, and how often you need updates — we'll come back with a defined scope and a quote.