Data Analyst CV: SQL, Python and Projects That Prove Impact

Data analyst roles attract enormous numbers of applications, and most of them look identical: a list of tools, a bootcamp certificate and a dashboard of a public dataset. What separates a shortlisted CV is not more tools — it is evidence that your analysis changed a decision.

Structure

1. Header: name, city, contact, plus links to GitHub, a portfolio or a published dashboard.
2. Profile: 3–4 lines on your domain, your main stack and the most valuable decision your work influenced.
3. Technical skills: grouped — languages, databases, BI tools, statistics and methods.
4. Experience: company, title, dates, data scale and 3–5 impact bullets.
5. Projects: two or three, only if they add something your jobs do not show.
6. Education and certifications.

Example profile

Data analyst with 4 years in retail e-commerce, working daily in SQL and Python on a warehouse of 40M+ order rows. Built the churn reporting that reallocated a six-figure retention budget and automated a reporting pack that saved the commercial team two days a month. Interested in analytics roles with direct product ownership.

Write impact, not tooling

Ran SQL queries is not an achievement — it is a precondition. Every bullet should contain a question, an analysis and a consequence:

• Built a cohort retention model in SQL and Python that identified a 19% drop in month-two retention after a pricing change, leading to a rollback.
• Replaced 14 manual Excel reports with a single Power BI model, cutting reporting time from two days to 20 minutes.
• Designed an A/B test for the checkout flow with 60,000 sessions per arm and defined the stopping rule with the product team.
• Rebuilt the revenue data model in dbt, reducing discrepancies between finance and product reporting to under 1%.
• Automated data quality checks that caught a broken tracking release within hours instead of weeks.

The technical block

• Languages: SQL (window functions, CTEs), Python (pandas, NumPy), R if you genuinely use it.
• Databases and warehouses: PostgreSQL, MySQL, BigQuery, Snowflake, Redshift.
• BI and visualisation: Power BI, Tableau, Looker, Metabase, Excel at an advanced level.
• Pipelines and versioning: dbt, Airflow, Git — strong differentiators for analytics engineering roles.
• Methods: A/B testing, regression, cohort and funnel analysis, forecasting, statistical significance.

List only what you could be questioned on. Fake depth in SQL is discovered in the first technical interview, usually within one question about joins or window functions.

No professional experience yet?

1. Use two or three real projects with messy data, not a cleaned tutorial dataset.
2. State the question first: which decision would this analysis support?
3. Show your SQL and your reasoning, not only the final chart.
4. Publish the code and link it. A reachable repository is worth more than a certificate.
5. Count analytical work from your previous job — reporting, pricing, operations analysis — as experience, because it is.

Nobody hires an analyst for the query. They hire for the decision the query made possible.