How to become a data analyst in 2026: a step-by-step roadmap
To become a data analyst, learn the tools in this order: Excel, SQL, basic statistics, one BI tool (Power BI or Tableau), then Python with pandas. Build 3 or 4 portfolio projects on real datasets and practise SQL interview questions. Studying 1 to 2 hours a day, most beginners are ready to apply in 4 to 6 months.
The 7 steps at a glance
- Learn Excel for analysis. Spend 2 weeks on formulas, XLOOKUP, pivot tables and charts. Excel is where most business data still lives.
- Learn SQL. Spend 4 weeks on SELECT, WHERE, GROUP BY, joins, subqueries, CTEs and window functions. SQL is the most tested skill in analyst interviews.
- Learn the statistics analysts use. Spend 2 to 3 weeks on averages, spread, distributions, sampling, confidence intervals, hypothesis tests and A/B tests.
- Learn one BI tool. Spend 2 to 3 weeks building dashboards in Power BI or Tableau, including data modelling and basic DAX or calculated fields.
- Learn Python for data analysis. Spend 4 weeks on Python basics, pandas, and charts with matplotlib or seaborn.
- Build a portfolio. Publish 3 or 4 end-to-end projects that start with a business question and end with a recommendation.
- Prepare for interviews. Practise SQL problems daily, rehearse explaining your projects, and apply to analyst and junior analyst roles.
What does a data analyst do?
A data analyst turns raw data into answers a business can act on. A typical week involves pulling data with SQL, cleaning it in Excel or Python, checking it with basic statistics, and presenting the result as a dashboard or a short written recommendation. The job is less about advanced algorithms and more about asking the right question, getting clean data and explaining what it means.
The skills you need, in order
| Step | Skill | Time (1 to 2 hrs a day) | You are ready when you can |
|---|---|---|---|
| 1 | Excel | 2 weeks | Clean a messy sheet, use XLOOKUP and build a pivot table and chart |
| 2 | SQL | 4 weeks | Join three tables, aggregate with GROUP BY and rank rows with window functions |
| 3 | Statistics | 2 to 3 weeks | Explain mean vs median, a p-value and how an A/B test works |
| 4 | Power BI or Tableau | 2 to 3 weeks | Build an interactive dashboard from a data model |
| 5 | Python (pandas) | 4 weeks | Load, clean, group and chart a dataset in a notebook |
| 6 | Portfolio | 3 to 4 weeks | Show 3 or 4 projects that end in a clear recommendation |
Step 1: Excel (weeks 1 and 2)
Start with Excel because it teaches you to think in rows, columns and aggregations without any code. Learn sorting and filtering, cleaning text and dates, IF, SUMIFS and COUNTIFS, XLOOKUP, pivot tables and charts. Finish by turning one messy sales sheet into a one-page summary.
Step 2: SQL (weeks 3 to 6)
SQL is the single most important analyst skill, and it is in almost every interview. Learn SELECT, WHERE, ORDER BY, GROUP BY and HAVING, then INNER and LEFT joins, subqueries, CTEs and window functions such as ROW_NUMBER, RANK and running totals. Practise on a free database such as PostgreSQL or SQLite. Our list of SQL interview questions for data analysts is a good checkpoint at the end of this step.
Step 3: Statistics (weeks 7 to 9)
You do not need a maths degree. You need to know when the mean misleads and the median does not, how spread is measured, what a distribution looks like, why sample size matters, and how to read a confidence interval, a p-value and an A/B test result. See the statistics concepts data analysts actually use.
Step 4: Power BI or Tableau (weeks 10 to 12)
Pick one BI tool and go deep rather than learning both lightly. Power BI is more common in companies that use Microsoft 365 and is widely asked for in Indian job listings. Tableau is popular in consulting and with US employers. Learn to import data, build a simple data model, write calculated measures and design a dashboard a manager can read in 30 seconds.
Step 5: Python for data analysis (weeks 13 to 16)
Python takes you past what Excel can handle. Learn the basics (variables, lists, dictionaries, loops and functions), then pandas for loading, cleaning, filtering, grouping and merging data, and matplotlib or seaborn for charts. Not sure whether to learn Python or SQL first? Read Excel vs SQL vs Python vs Power BI.
Step 6: Build a portfolio (weeks 17 to 20)
Recruiters hire analysts who can show finished work. Build 3 or 4 projects, each one following the same shape: a business question, the data source, the cleaning you did, the analysis, one dashboard or chart, and a recommendation in plain language. Good starter topics are sales performance, customer churn, an A/B test and an HR attrition analysis. Put them on GitHub or a free portfolio site and link them on your CV and LinkedIn.
Step 7: Prepare for interviews
Most analyst interviews have three parts: a SQL test, a case or business question, and a walk-through of your projects. Solve one or two SQL problems a day, practise explaining a project in two minutes, and prepare a short story for "tell me about a time data changed a decision".
Common mistakes to avoid
- Learning many tools shallowly instead of SQL and one BI tool deeply.
- Watching courses without building anything you can show.
- Skipping statistics, then being unable to explain whether a result is meaningful.
- Portfolio projects that end with a chart but no recommendation.
Frequently asked questions
How long does it take to become a data analyst?
Can I become a data analyst without a degree?
Do data analysts need to know coding?
Should I learn Power BI or Tableau first?
What is the first skill a data analyst should learn?
Learn it from handwritten notes
Data Analytics: The Handwritten Notes: 116 illustrated pages, 26 chapters, drawn diagrams and worked examples. ₹199 in India, US$9 elsewhere.
See what is inside