The 6-month data analyst roadmap: Month 1 — Excel & SQL foundations. Month 2 — Advanced SQL & data cleaning. Month 3 — Python (pandas, matplotlib). Month 4 — Visualization (Tableau or Power BI). Month 5 — Statistics & business context. Month 6 — Portfolio polish & job search. Each month has a specific deliverable that becomes part of your portfolio. The order matters — SQL first, Python third, not the other way around.
This article was researched and written by the Careery team — that helps land higher-paying jobs faster than ever! Learn more about Careery →
How long does it take to become a data analyst from scratch?
With a structured plan: 6 months studying 15–20 hours per week (part-time), or 3 months at 35–40 hours per week (full-time). The critical path is SQL (6 weeks) → Excel (2 weeks) → BI tool (4 weeks) → Python (4 weeks) → statistics (3 weeks) → portfolio polish (3 weeks). Rushing through any stage creates gaps that surface during interviews.
What should I learn first to become a data analyst?
SQL — not Python, not statistics, not machine learning. SQL appears in 90%+ of data analyst job postings and is used daily in virtually every analyst role. Start with SELECT, WHERE, GROUP BY, JOINs, then advance to window functions and CTEs. You can build your first portfolio project after just 4–6 weeks of SQL study.
Can I become a data analyst in 3 months?
Yes, if studying full-time (35–40 hours per week) and following a structured curriculum. The minimum viable skill set — SQL + Excel + one BI tool — can be learned in 8–10 weeks. The remaining time goes to portfolio projects and job search preparation. Part-time learners should plan for 6 months.
Most people who try to become data analysts don't fail because the material is too hard. They fail because they study the wrong things in the wrong order. Spending three months on Python before touching SQL. Taking a machine learning course before understanding basic statistics. Building a portfolio with cleaned Kaggle datasets that every other beginner uses.
This roadmap fixes that. Each month builds on the previous one, and every month produces a tangible deliverable. By Month 6, the portfolio is built, the skills are proven, and the job search is strategic — not desperate.
| Month | Focus | Key Deliverable | Hours/Week |
|---|---|---|---|
| Month 1 | Excel & SQL Foundations | First SQL portfolio project (data exploration) | 15–20 |
| Month 2 | Advanced SQL & Data Cleaning | Complex SQL project with multiple JOINs and CTEs | 15–20 |
| Month 3 | Python (pandas, matplotlib) | Python data cleaning and analysis notebook | 15–20 |
| Month 4 | Visualization (Tableau or Power BI) | Published interactive dashboard on Tableau Public | 15–20 |
| Month 5 | Statistics & Business Context | End-to-end analysis with statistical insights and recommendations | 15–20 |
| Month 6 | Portfolio Polish & Job Search | Polished portfolio, optimized resume, active applications | 20–25 |
This roadmap covers the learning path. For the full picture — including education options, career progression, and salary expectations — see How to Become a Data Analyst in 2026.
Six months, six deliverables. Each month produces a portfolio piece. The order is non-negotiable: SQL and Excel first because they're required for every role, Python and visualization next because they make you competitive, statistics and business context last because they make you senior-track.
Here's what each month looks like in detail.
Most beginners want to skip Excel and jump straight to Python. That's a mistake. Excel is still the language that non-technical stakeholders speak, and SQL is the tool every data analyst uses daily. These two skills alone can get someone hired at a company that needs basic reporting help.
Weeks 1–2: Excel & Google Sheets
What to learn:
- Pivot tables — the single most important Excel skill for analysts
- VLOOKUP and INDEX-MATCH for data lookups across sheets
- Conditional formatting to highlight outliers and trends
- Basic charts (bar, line, scatter) for quick exploratory analysis
- Data validation and cleaning (removing duplicates, handling blanks, standardizing formats)
Resources (free):
- Google Sheets official training (free, covers 80% of what analysts need)
- Excel practice datasets from Kaggle (search "sales data" or "customer data")
Weekly goal: Build one pivot table analysis per week using a real dataset. By the end of Week 2, the ability to go from raw CSV to pivot table summary in under 10 minutes should be automatic.
Weeks 3–4: SQL Foundations
What to learn:
- SELECT, FROM, WHERE, ORDER BY — the basics of querying
- GROUP BY and aggregate functions (COUNT, SUM, AVG, MIN, MAX)
- JOINs (INNER, LEFT, RIGHT) — the most important SQL concept for analysts
- Filtering with HAVING, IN, BETWEEN, LIKE
- Basic subqueries
Resources (free):
- SQLBolt (interactive, browser-based — no setup required)
- Mode Analytics SQL tutorial (uses real datasets)
- W3Schools SQL reference (for quick syntax lookups)
Month 1 deliverable: A SQL project that explores a public dataset (e.g., NYC taxi data, Census data). Write 10–15 queries that answer business questions, document the queries in a GitHub README, and explain the findings.
Month 1 establishes the foundation: Excel for stakeholder communication and quick analysis, SQL for querying databases. The deliverable — a documented SQL exploration project — becomes the first portfolio piece.
With the basics solid, Month 2 moves to the SQL skills that separate junior applicants from hirable candidates.
Basic SQL gets you through a recruiter screen. Advanced SQL gets you through the technical interview. The difference between "knows SQL" and "is good at SQL" comes down to three concepts: window functions, CTEs, and data cleaning patterns.
Weeks 5–6: Advanced SQL
What to learn:
- Window functions: ROW_NUMBER, RANK, DENSE_RANK, LAG, LEAD, running totals
- Common Table Expressions (CTEs) for readable, modular queries
- CASE statements for conditional logic and data categorization
- Date functions for time-series analysis (DATE_TRUNC, EXTRACT, DATEDIFF)
- Performance basics: understanding indexes, avoiding SELECT *
Resources:
- SQL for Data Analytics by Jun Shan (O'Reilly, 2025) — the best analyst-specific SQL reference
- LeetCode SQL problems (Easy and Medium difficulty)
- StrataScratch — real interview SQL questions from companies
Weeks 7–8: Data Cleaning in SQL & Spreadsheets
What to learn:
- Handling NULL values (COALESCE, NULLIF, IS NULL patterns)
- Deduplication techniques (ROW_NUMBER for identifying duplicates)
- String cleaning (TRIM, UPPER, LOWER, REPLACE, REGEXP)
- Data type conversions and format standardization
- Joining messy datasets with inconsistent keys
Month 2 deliverable: A complex SQL project that joins 3+ tables, uses window functions and CTEs, and answers a multi-part business question. Example: "Analyze customer retention by cohort, calculate month-over-month churn, and identify the top 3 factors correlated with churn." Publish the queries and findings on GitHub.
Window functions and CTEs are the two SQL skills that most entry-level candidates lack. Mastering them in Month 2 puts you ahead of 70% of applicants. The deliverable should demonstrate querying multiple tables, handling messy data, and answering layered business questions.
SQL and Excel handle 60% of data analyst work. Python handles the next 25%.
Python for data analysis is not the same as Python for software engineering. Analysts don't need to understand object-oriented programming, web frameworks, or algorithms. They need pandas for data manipulation, matplotlib for visualization, and Jupyter notebooks for documentation.
Weeks 9–10: Python & pandas Fundamentals
What to learn:
- Python basics: variables, loops, conditionals, functions (just enough to use pandas)
- pandas: reading CSVs, DataFrames, filtering, groupby, merge, pivot tables
- Data cleaning in pandas: handling NaN values, string operations, type conversions
- Jupyter Notebooks for combining code, output, and documentation
Resources:
- Python for Data Analysis by Wes McKinney (O'Reilly, 3rd edition) — written by the creator of pandas
- Kaggle's free Python and pandas micro-courses (browser-based, no setup)
Weeks 11–12: Visualization & Automation
What to learn:
- matplotlib basics: line charts, bar charts, scatter plots, histograms
- seaborn for statistical visualizations (heatmaps, pair plots, distribution plots)
- Basic automation: writing Python scripts to clean and transform data that arrives regularly
- Exporting results to CSV/Excel for stakeholder consumption
Month 3 deliverable: A Jupyter Notebook that imports a messy real-world dataset, cleans it using pandas, performs exploratory analysis, and produces 5–7 visualizations with written interpretations. Publish on GitHub with a README explaining the analysis.
- Spending 3 weeks on Python syntax before touching pandas — analysts need pandas skills, not CS fundamentals
- Trying to learn scikit-learn or machine learning — that's data science, not data analysis
- Building projects in .py files instead of Jupyter Notebooks — notebooks are the analyst's presentation format
- Perfecting code style instead of producing analysis — hiring managers evaluate insights, not PEP 8 compliance
Python for analysts means pandas + matplotlib + Jupyter Notebooks. Python for Data Analysis by Wes McKinney is the definitive reference. Spend 70% of Month 3 on pandas (data manipulation) and 30% on matplotlib (visualization). Skip everything else.
Python produces the analysis. Month 4 turns that analysis into something stakeholders actually want to look at.
Visualization is where technical skill meets communication. A Tableau dashboard or Power BI report is often the primary deliverable a data analyst produces — and it's what stakeholders judge competence by.
Weeks 13–14: Tableau or Power BI Fundamentals
What to learn (pick one tool):
- Connecting to data sources (CSV, Excel, databases)
- Building core chart types: bar, line, scatter, heat map, treemap
- Filters, parameters, and calculated fields
- Dashboard layout and design principles
- Interactivity: tooltips, filter actions, drill-downs
How to choose: Tableau is more common at tech companies and startups. Power BI dominates at Microsoft ecosystem companies and enterprises. Check job postings in the target market and learn whichever appears more frequently. Learning one makes learning the other straightforward.
Weeks 15–16: Dashboard Design & Storytelling
What to learn:
- Cole Nussbaumer Knaflic's Storytelling with Data principles: clutter reduction, strategic color use, narrative structure
- Dashboard hierarchy: KPIs at top, trends in middle, details on demand
- Designing for the audience: executive dashboards vs. operational dashboards
- Publishing to Tableau Public (free) for portfolio visibility
Month 4 deliverable: An interactive dashboard published on Tableau Public (or Power BI shared link) that tells a complete data story. Include: a clear title, 4–6 charts, filter controls, and a text summary of key findings. Use a dataset different from previous months.
For the full technical breakdown of each tool and what proficiency looks like at each level, see Data Analyst Skills & Tools.
Visualization is the skill stakeholders evaluate directly. A published Tableau dashboard is the most impactful single portfolio piece. Storytelling with Data by Cole Nussbaumer Knaflic (Wiley, 2015) is the essential reference — read it during Month 4 and apply its principles to every chart built from this point forward.
With tools mastered, Month 5 adds the judgment layer that separates competent analysts from great ones.
Statistics makes analysis trustworthy. Business context makes it useful. Month 5 combines both — learning enough statistics to validate findings and enough business thinking to make recommendations that actually get implemented.
Weeks 17–18: Practical Statistics
What to learn:
- Descriptive statistics: mean, median, mode, standard deviation, percentiles
- Distributions: normal, skewed, bimodal — and why they matter
- Correlation vs. causation (the most important statistical concept for analysts)
- Hypothesis testing basics: p-values, confidence intervals, A/B test interpretation
- Statistical significance: when a finding is real vs. when it's noise
Resources:
- Naked Statistics by Charles Wheelan — the most accessible statistics book for non-mathematicians
- Khan Academy Statistics & Probability (free, structured)
Weeks 19–20: Business Context & Communication
What to learn:
- Reading financial statements (income statement, balance sheet basics)
- Understanding KPIs by function: marketing (CAC, LTV, conversion), product (DAU, retention, churn), finance (revenue, margin, burn rate)
- Structuring a data presentation: situation → analysis → recommendation → impact
- Presenting to non-technical audiences: lead with the "so what," not the methodology
Month 5 deliverable: An end-to-end analysis project that includes data cleaning, statistical analysis, visualization, and a written recommendation. Structure it like a business memo: executive summary, methodology, findings, recommendations, and appendix with technical details.
Statistics validates findings. Business context makes those findings actionable. The Month 5 deliverable should read like a consulting report, not a homework assignment — clear recommendations backed by data, written for someone who doesn't know SQL.
Five months of skills. Month 6 turns them into a job.
The difference between "learning data analytics" and "getting hired as a data analyst" is packaging and strategy. Month 6 is entirely about converting skills into employment.
Weeks 21–22: Portfolio Polish
Actions:
- Audit all portfolio projects: clean READMEs, consistent formatting, clear business context
- Ensure every project has: a problem statement, methodology, key findings, and recommendations
- Create a portfolio landing page (GitHub profile README or personal site)
- Publish all Tableau dashboards on Tableau Public
- Record 2-minute Loom walkthroughs for top 2 projects
Weeks 23–24: Job Search Execution
Actions:
- Build a tailored resume using the [action + analysis + tool + impact] bullet formula
- Customize resume for 3–4 role categories (startup, enterprise, industry-specific)
- Apply to 50–75 roles over 4 weeks, weighted toward mid-size companies
- Practice SQL interview problems daily (15–20 minutes on StrataScratch or LeetCode)
- Prepare a 2-minute career story and 3 case study walkthroughs
For specific project ideas, README templates, and examples of portfolios that got people hired, see Data Analyst Portfolio Projects That Actually Get You Hired.
Month 6 converts five months of skill-building into a job. Polish the portfolio, build a tailored resume, and apply strategically. Quantity matters — 50–75 applications over 4 weeks, customized by role category, not individual posting.
- 01Month 1: Excel & SQL foundations — build a SQL exploration project as the first portfolio piece
- 02Month 2: Advanced SQL (window functions, CTEs) & data cleaning — demonstrate complex querying ability
- 03Month 3: Python (pandas + matplotlib) — add data manipulation and visualization to the toolkit
- 04Month 4: Tableau or Power BI — publish an interactive dashboard that tells a data story
- 05Month 5: Statistics & business context — produce an end-to-end analysis with clear recommendations
- 06Month 6: Portfolio polish & job search — apply to 50–75 roles with a tailored resume and published portfolio
Can I follow this roadmap while working full-time?
Yes. The roadmap assumes 15–20 hours per week, which is manageable alongside a full-time job — typically 2–3 hours on weekday evenings and 4–5 hours on weekends. The timeline extends to 6 months (vs. 3 months for full-time study). Consistency matters more than intensity.
Should I learn Tableau or Power BI?
Check job postings in your target market. Tableau is more common at tech companies and startups; Power BI dominates in Microsoft ecosystem companies and large enterprises. If unsure, start with Tableau — it has a free public version for building a portfolio and is slightly more common in job postings overall.
Do I need to pay for courses?
No. Every skill in this roadmap can be learned with free resources: SQLBolt for SQL, Kaggle micro-courses for Python, Tableau Public for visualization, and Khan Academy for statistics. Paid options like the Google Data Analytics Certificate ($49/month on Coursera) are worthwhile for the structured curriculum and the certification credential, but they're optional.
What if I already know Excel?
Skip the Excel portion of Month 1 and start SQL immediately. Use the extra 2 weeks to go deeper on SQL fundamentals or start Month 2's advanced SQL topics earlier. Existing Excel proficiency is a significant advantage — it means you already think about data in tabular structures.
How do I know when I'm ready to apply?
Use the Job-Readiness Assessment checklist in this article. If you can check 7+ of the 10 items, start applying. You don't need to be perfect — entry-level roles expect you to learn on the job. The biggest mistake is waiting until you feel '100% ready,' which usually means waiting too long.
Prepared by Careery Team
Researching Job Market & Building AI Tools for careerists · since December 2020
- 01Python for Data Analysis: Data Wrangling with pandas, NumPy, and Jupyter — Wes McKinney (2022 (3rd edition))
- 02Storytelling with Data: A Data Visualization Guide for Business Professionals — Cole Nussbaumer Knaflic (2015)
- 03Occupational Outlook Handbook: Data Analysts and Scientists — Bureau of Labor Statistics (2025)
- 04Naked Statistics: Stripping the Dread from the Data — Charles Wheelan (2013)