Quality Guidelines

High-quality assets drive repeat buyers, positive ratings, and sustainable revenue. These are the standards Spartera expects from all marketplace assets — both at publish time and ongoing.

Bad data = bad insights. No amount of good SQL fixes a corrupted or stale data source.

Pre-Publication Requirements

Every asset must meet these before going live:

Technical

Preview execution completes successfully with no errors
Execution time under 2 seconds for standard queries (30 seconds max for complex analytics)
Output schema is documented — buyers need to know what columns/fields to expect
Error conditions return clear messages, not raw database errors
Null and edge cases are handled gracefully (no division-by-zero crashes, no unhandled empty results)

Listing Quality

Asset name is specific and searchable ("30-Day Churn Rate" not "Customer Metric")
Description explains: what it returns, what data powers it, how fresh the data is, and what business decision it enables
Category and tags are set accurately
Price is set and reflects the value delivered

Data Quality Standards

Accuracy

Aggregations and calculations are mathematically correct
Data transformations preserve the meaning of the source data
Joins maintain referential integrity
Outliers are handled appropriately (not silently dropped or ignored)

Freshness

Be transparent about data recency. Your listing should state:

How often the underlying data is updated ("updated daily at 2 AM UTC")
The data's historical coverage ("covers the trailing 12 months")
How buyers will know if data is stale

Update Cadence	Acceptable Use Case
Real-time (seconds)	Live trading signals, active monitoring
Near real-time (minutes)	Operational dashboards, alert systems
Daily	Most business analytics, trend analysis
Weekly/monthly	Strategic reporting, benchmarking

Source Reliability

Use validated, production-quality data sources
Avoid test data, placeholder records, or development databases
If your source data has known gaps, document them

Performance Standards

Query Type	Target	Maximum
Simple aggregation	< 500ms	2s
Multi-table join	< 1s	5s
Complex analytics / ML	< 5s	30s

If your asset regularly exceeds these targets, optimize before publishing:

Add appropriate indexes to your source tables
Avoid SELECT * — specify only needed columns
Pre-aggregate data in a summary table if the raw query is too slow
Use LIMIT on large result sets

Documentation Standards

Required in Every Asset Description

What it returns — specific output fields and their meaning
Data source — what database/table/system powers it and who maintains it
Freshness — how current the data is
Use case — what business question or decision this answers

Ongoing Standards (Post-Publish)

Quality requirements don't stop at publish. Spartera monitors live assets for:

Error rate — assets exceeding 5% error rate on production executions will be flagged
Response time degradation — if execution time degrades significantly, you'll be notified
Broken connections — if your database connection becomes invalid, the asset is auto-suspended until fixed
Buyer complaints — repeated reports of inaccurate results trigger a review

Versioning Policy

Non-breaking changes (query optimization, minor filter changes): patch update, same version
Breaking changes (new/removed output columns, renamed fields): must bump major version and notify existing buyers with at least 30 days notice

Enforcement

Assets that consistently fail quality standards will be:

Flagged and the seller notified
Suspended from the marketplace if not fixed within 14 days
Removed if quality issues persist

Spartera's goal is not to penalize sellers — it's to protect buyer trust. If you're struggling with a quality issue, contact [email protected] before your asset gets flagged.

Testing and Validation — How to test before publishing
Publishing to Marketplace — Full publish guide
Data Sources — Supported connections

Quality Guidelines