Data Quality Scorecards measure how clean your Golden DB is: completeness per field, accuracy per source, freshness per record. Reviewed weekly, they reveal where data quality is degrading and which sources to trust.
What the scorecard measures
Four dimensions per record class:
- Completeness — what % of records have each field filled.
- Accuracy — estimated correctness based on cross-source agreement.
- Freshness — how recently each field was updated.
- Consistency — whether duplicate-resolution found expected duplicates.
Where to find it
Data → Quality → Scorecard. The top view shows aggregate scores per record class (Accounts, Contacts, Deals). Drill into each for per-field detail and per-source quality.
Reading the per-field scorecard
Each field shows: % filled, source distribution (which provider filled how many), and an accuracy estimate (where cross-source agreement is high vs disputed). Low fill rates indicate gaps in your provider mix; high source-concentration indicates you're dependent on one provider for that field.
Per-source quality
Each connected data source has a quality score per field. High score = this source's values agree with other sources and downstream signals (e.g. emails get replies). Low score = this source's values get overridden by other sources or fail downstream validation. Use these scores to refine your waterfall enrichment cascade.
Weekly review cadence
Open the scorecard weekly and spot the changes: scores trending down on any field, freshness aging on any segment, new gaps from a provider connectivity issue. Most issues are easy to fix if caught early.
