Dashboards sit on top of a capture layer (DB / warehouse / SaaS APIs / spreadsheets); data quality below the surface determines trust — abandonment frequently stems from numbers that don't reconcile, not from the visualization

Claim: Dashboards connect to operational databases, data warehouses (Snowflake, BigQuery), SaaS APIs, and spreadsheets. The capture layer + data quality are the foundation: abandonment frequently stems from data the user does not trust (numbers that do not reconcile with the source system), not from visualization choices.

Source: Practitioner / vendor synthesis (consistent across BI vendor literature).

Confidence: Industry-consensus.

Why this matters for Candid: The build-the-dashboard question is downstream of "is the capture layer right?" Same logic as the data-tools brief (R2 — Build only on data you already own — transaction history, CRM, scheduling, no-show patterns; that is the only category with native defensibility) — you cannot dashboard data you have not captured cleanly.