analyst_skill

Overview

This skill helps extract actionable insights from data using SQL, visualizations, and clear written summaries. It emphasizes decision-focused analysis, reproducible workflows, and delivering recommendations that stakeholders can act on. Use it to move from raw data to confident decisions quickly.

How this skill works

The skill guides you through framing a decision, validating data quality, and constructing efficient SQL queries with readable patterns like CTEs and window functions. It recommends analysis approaches (cohorts, segmentation, time-series) and matches visualization types to messages. Finally, it produces concise findings, confidence levels, and practical recommendations for stakeholders.

When to use it

• You need to answer a business decision, not just explore data.
• Preparing a stakeholder-ready summary from a raw dataset.
• Building repeatable SQL queries and automated reports.
• Diagnosing data quality issues before modeling or reporting.
• Designing charts that communicate one clear message.

Best practices

• Start by clarifying the decision and hypothesis before querying.
• Validate row counts, date ranges, nulls, and duplicates first.
• Prefer readable SQL (CTEs, aggregates before joins) over clever hacks.
• Use cohorts and segmentation to surface behavior hidden by aggregates.
• Lead with the insight, then show evidence and recommend next steps.

Example use cases

• Assessing whether a recent product change impacted retention using cohort analysis.
• Building a weekly revenue report that aggregates before joining for speed.
• Diagnosing a sudden drop in funnel conversion by validating source definitions and duplicates.
• Designing a dashboard with one-message charts and table for exact figures.
• Automating a monthly churn calculation with reproducible SQL and version control.