Dashboards, KPIs, and storytelling.
2 topics
BI analysts query data warehouses to produce insights. Master SQL: complex joins, window functions, CTEs, date manipulation, and performance optimization. Understand data warehouse concepts: star/snowflake schemas, fact vs dimension tables, slowly changing dimensions, and grain. Tools: Snowflake, BigQuery, Redshift. dbt for transformations.
2 resources
Define metrics that drive business decisions. Frameworks: North Star Metric, AARRR (Acquisition, Activation, Retention, Revenue, Referral), and balanced scorecards. Design metric hierarchies: company-level OKRs → team KPIs → operational metrics. Understand leading vs lagging indicators. Ensure metrics are: measurable, actionable, relevant, and time-bound. Avoid vanity metrics.
Build dashboards that drive action. Tools: Tableau, Power BI, Looker, Metabase. Principles: one dashboard per audience/purpose, most important metric at top-left, minimal chart types per dashboard, consistent colors, and interactive filters. Design hierarchy: executive dashboards (5-7 KPIs), operational dashboards (detailed metrics), and self-serve exploration. Maintain a single source of truth with a semantic layer.
Communicate findings that drive decisions. Structure: context (business question) → analysis (methodology) → findings (key insights with visuals) → recommendations (specific actions). Use the pyramid principle: lead with conclusion, support with evidence. Tailor to audience: execs want decisions, managers want trends, analysts want methodology. Present uncertainty honestly with confidence intervals.