Skip to main content

Product intelligence

Product intelligence is the layer that turns raw catalog rows into decisions: what to fix, what to publish, what will fail on a channel, and what AI should work on next.

Lead with agent findings from the Agent Control Center; use completeness scores as publish gates — they measure different things. See Catalog quality — findings not fill rates.

Signals

SignalWhat it tells you
Agent insightsOpen findings from specialist agents — SEO, taxonomy, compliance, syndication
Catalog Quality scoreRollup of open insights + pending approvals (ACC)
Approval queueAgent-proposed writes awaiting human review
Completeness score% of required attributes filled per family/channel — publish gate
Enrichment queueSKUs below threshold needing AI or manual work
Export job itemsPer-SKU syndication success/failure
Task backlogWorkflow approvals blocking go-live
Dashboard KPIsVolume, activity, assignments (MCP: get_dashboard_stats)

Intelligence in action

APIs and tools

OperationRESTMCP
Product completenessProduct APIget_product_completeness, analyze_product_completeness
Find gapsSearch / reportsfind_products_needing_enrichment
Compare SKUscompare_products
RecalculateProduct APIrecalculate_product_completeness

Why this beats spreadsheets

Spreadsheets do not know channel rules or family schemas. CataZenta intelligence is computed from the same model that powers Amazon, Shopify, and your agents — one source of truth, many outputs.

Catalog quality · Data model · Completeness guide