What a Data Analyst does
A Data Analyst turns raw data into useful answers that guide decisions. You’ll ask clarifying questions, pull data (often with SQL), clean it, analyze trends, and present clear takeaways to stakeholders. Your work connects business problems to measurable results.
See a sample week
- Monday: Clarify a stakeholder question about churn. Define the metric and time window.
- Tuesday: Write SQL to extract cohorts; check data quality; create first cut of results.
- Wednesday: Explore patterns, segment by plan/region; create a concise chart.
- Thursday: Review with product manager; refine and add a simple forecast.
- Friday: Share a 1-page summary with decision and next steps; log learnings.
Typical deliverables
- Clean datasets or views ready for analysis.
- Dashboards with KPIs and trends.
- One-pagers or slide summaries with insights and recommendations.
- Experiment readouts (A/B test results, lift, confidence).
- Ad-hoc analyses answering specific business questions.
Where you can work
- Industries: tech, e-commerce, fintech, healthcare, education, logistics, marketing.
- Teams: product analytics, marketing analytics, operations, risk/fraud, finance FP&A.
- Company sizes: startups (generalist), mid-size (focused domains), enterprise (specialized analysts).
Salary expectations
Approximate total annual compensation in USD (Varies by country/company; treat as rough ranges.)
- Junior: $50k–$80k
- Mid-level: $80k–$120k
- Senior: $110k–$160k+
Hiring expectations by level
Junior / Entry
- Writes basic SQL (SELECT, WHERE, GROUP BY).
- Builds simple charts and explains trends clearly.
- Understands core metrics and basic statistics (averages, medians, percentages).
- Needs guidance on scoping and data validation.
Mid-level
- Owns projects end-to-end: scoping, data, analysis, story, recommendation.
- Comfortable with joins, window functions, and quality checks.
- Designs useful dashboards; collaborates well with product/marketing/ops.
- Explains trade-offs and assumptions; proposes next experiments.
Senior
- Leads ambiguous problems and aligns stakeholders on the right metrics.
- Builds reliable datasets; mentors others on analytic rigor and communication.
- Shapes experimentation strategy and standards for the team.
- Drives measurable business outcomes with clear decision frameworks.
Who this is for
- People who enjoy structured problem-solving and clear communication.
- Those comfortable with numbers and curious about how products grow.
- Career switchers seeking a business-impact role with technical fundamentals.
Prerequisites
- Comfort with basic arithmetic, percentages, and averages.
- Willingness to learn SQL and interpret charts.
- Attention to detail and a habit of validating data.
Skill map and your first skill
Start with SQL. It’s the backbone for most analyst tasks: extracting data, creating aggregates, and preparing tables for dashboards and experiments. You can add spreadsheets and visualization tools later, but SQL unlocks real datasets quickly.
Mini task: Write your first query
Imagine a table orders(user_id, created_at, amount, status). Goal: find total revenue from completed orders last 30 days.
SELECT
SUM(amount) AS revenue_30d
FROM orders
WHERE status = 'completed'
AND created_at >= CURRENT_DATE - INTERVAL '30 days';
Variation: Group by day and compute average order value per day.
Learning path
Practical portfolio projects
- E-commerce funnel analysis: From sessions to purchase. Outcome: a funnel chart and top 3 drivers of drop-off.
- Cohort retention: Define weekly cohorts and day-7 retention. Outcome: a table/heatmap and 2 actionable recommendations.
- Marketing channel ROI: Compare CAC and LTV by channel. Outcome: a 1-page decision memo.
- Churn drivers: Segment cancellations by plan, tenure, and region. Outcome: prioritized list of interventions.
- Operational dashboard: On-time delivery rate and SLA breaches. Outcome: a daily KPI dashboard with alerts criteria.
Interview preparation checklist
- Clarify the business question, success metric, and time horizon before querying.
- Sketch table relationships and sample rows to avoid wrong joins.
- Explain trade-offs: metric definitions, sampling, and potential biases.
- Practice SQL on realistic schemas (orders, events, users, subscriptions).
- Tell concise stories: context → method → finding → impact → next step.
- Prepare one 3–5 minute walkthrough for each portfolio project.
- Have a data-quality checklist: duplicates, missing values, timezones, status flags.
Common mistakes and how to avoid them
Jumping into SQL without scoping
Solution: Confirm the decision to be made, the owner, and the exact metric definition before writing queries.
Wrong joins and double-counting
Solution: Validate row counts after joins; use DISTINCT or window functions when appropriate; sample a few user histories.
No data-quality checks
Solution: Check nulls, duplicates, and timestamp ranges. Compare to a known baseline or prior periods.
Overcomplicated charts
Solution: Start with one clear chart per insight. Label units and time windows. Add context and the recommendation.
Next steps
Pick a skill to start. Begin with SQL in the Skills section below, then come back for the exam when you can reliably compute core metrics.