data-analyst_skill

Overview

This skill teaches SQL for data analysis focused on exploratory analysis, advanced aggregations, statistical functions, outlier detection, and actionable business insights. It includes 50+ real-world analytics queries and patterns to help you profile data, validate quality, run cohort and trend analyses, and extract insights for decisions.

How this skill works

The skill provides ready-to-run SQL patterns that inspect table structure, value distributions, missingness, and statistical summaries. It implements window functions, percentiles, IQR and z-score outlier detection, cohort retention, period-over-period comparisons, correlation estimators, and moving averages to reveal trends and anomalies. Examples are written to be adapted across common SQL dialects and production workflows.

When to use it

• Perform initial exploratory data analysis on a new dataset to assess quality and distributions
• Detect and investigate outliers before feeding data to models or reports
• Compare sales or KPIs across periods and compute growth and percent changes
• Build cohort retention and user behavior analyses for product or marketing teams
• Validate data quality rules and find duplicates or impossible values

Best practices

• Start with data profiling queries to understand missingness and value ranges before deep analysis
• Use window functions and grouped percentiles for robust summary stats by segment
• Apply both IQR and z-score methods to cross-check outlier detection results
• Materialize expensive intermediate results (CTEs or temp tables) when reusing stats across queries
• Adapt date-truncation and percentile functions to your SQL dialect for consistent results

Example use cases

• Generate department-level salary summaries and detect compensation anomalies
• Compute 7- and 30-day moving averages for daily revenue monitoring
• Run cohort analysis to measure monthly user retention after acquisition campaigns
• Identify invalid records such as future hire dates or malformed emails for data cleaning
• Produce period-over-period sales reports with absolute and percentage change