Global Data Science Job Market Analysis (PostgreSQL)

A pure-SQL analysis of 3,925 global data science job listings, built entirely in PostgreSQL using window functions, CTEs, array parsing, and statistically-aware aggregation to uncover salary patterns across countries, experience levels, and job specializations.

Project Summary

This project answers a core career-strategy question for data professionals: where does experience actually pay off, which roles command the biggest premiums, and how do country-level salary averages hide or distort the real picture? Every query in this project was written and debugged directly in PostgreSQL — no Python, no spreadsheet tool — using the dataset's full 3,925 rows across company location, job title, job category, experience level, and salary in USD.

A key discipline applied throughout: every aggregation query carries a minimum sample size filter (HAVING COUNT(*) >= 5 or its window-function equivalent) to prevent single-listing outliers from distorting country or role-level averages — a statistical safeguard many beginner SQL projects skip entirely.

Key Insights

• Senior talent monopolizes the market. Senior-level roles account for 67.2% of all active listings (2,638 of 3,925 rows) — companies are overwhelmingly favoring proven, day-one execution over entry-level training.
• The experience premium is real and steep. Moving from an Entry-level baseline ($77,953 avg) to Senior-level ($147,571 avg) delivers an 89.3% salary jump — one of the clearest career ROI signals in the dataset.
• Executive roles are rare but elite. Executive and leadership positions make up just 3.4% of global listings, yet command the highest average salary baseline at $198,568.
• Local specialization beats regional baselines. While a standard Data Analyst in Australia earns $165,000–$171,000, a specialized ML Engineer in that same market commands up to $300,000 — a $129,000 premium over the local Data Analyst benchmark.
• Country-wide averages mask the real top earners. Australia's national salary baseline sits at $123,648, but window-function benchmarking reveals specialized ML Engineering roles peaking at $300,000 — a $176,352 premium over the national average, more than double typical regional earnings. The same pattern holds in emerging hubs: a Principal Machine Learning Engineer in India secures $160,000, well above that market's general baseline.

Tools and SQL Techniques Used

• PostgreSQL via pgAdmin
• Data cleaning: NULL audits, duplicate detection, text normalization with CASE + ILIKE
• Feature engineering: custom experience-band categorization via ALTER TABLE + UPDATE
• Aggregation: GROUP BY, HAVING, COUNT(*) FILTER
• Window functions: RANK(), PERCENT_RANK(), PARTITION BY, windowed AVG() and COUNT()
• CTEs (Common Table Expressions) for multi-step ranking logic
• Statistical safeguarding: minimum sample-size thresholds applied consistently across every aggregation to prevent outlier distortion

Dataset Source

Global Data Science Job Market dataset, sourced from Kaggle. 3,925 rows covering job title, job category, experience level, company location, and salary in USD.

Repository

https://github.com/PranavRoy07/Global-Data-Science-Job-Market-Analysis-PostgreSQL-

Author

Pranav — Aspiring Data Analyst | BBACA Graduate