Six months at 10 to 12 hours a week gets a beginner from zero to job-ready, in this order: Excel, then SQL, then a BI tool, then Python, then statistics, then a portfolio and applications. Most roadmaps you will find are lists of topics with no sequence and no stopping rule, which is why people spend four months on Python and still cannot answer a business question.
This one gives each month a single focus, one thing you build, and a checkpoint that tells you whether you are allowed to move on. Skipping the checkpoint is how people end up six months in with six half-finished courses.
Settle two things before month one
How many hours a week you actually have. Not the number you would like. Ten to twelve hours is the realistic figure for someone working full time, and the plan below assumes it. At twenty hours a week you can compress this into three to four months. At five hours a week the plan takes a year, and you should know that going in rather than quitting in month three feeling like you failed.
Which job you are aiming at. “Data analyst” covers a business analyst seat in a bank, a reporting analyst at an IT services firm and a product analyst at a startup. They share a core stack, so you do not need to decide today, but by month five your projects should look like the job you want. Our guide to data analyst qualifications in India covers what employers screen for before any of this starts.
The six months at a glance
| Month | Focus | What you build | Checkpoint before moving on |
|---|---|---|---|
| 1 | Excel and how business data is shaped | A cleaned sales workbook with a pivot summary | You can build a pivot table and an INDEX MATCH without looking them up |
| 2 | SQL | 40 queries against a multi-table database | You can write a three-table join with a GROUP BY unaided |
| 3 | Power BI or Tableau | One dashboard that answers a stated question | A stranger can read your dashboard and tell you the finding |
| 4 | Python for analysis | A notebook that cleans a messy CSV end to end | You can load, clean, group and plot without a tutorial open |
| 5 | Statistics and a second project | A project with a written conclusion, not just charts | You can explain your result and its limits out loud in two minutes |
| 6 | Portfolio, resume, applications | Three finished projects and 50 applications sent | You have had at least one first-round interview |
Month 1: Excel, because that is where the data still lives
Start with Excel and give it a full month. It is the tool Indian companies actually run their reporting on, interviews still include a pivot or lookup exercise, and it teaches you the shape of business data faster than any programming language will.
What to cover: cleaning and deduplication, text-to-columns, VLOOKUP and INDEX MATCH, pivot tables, conditional formatting, and the date functions that make Indian financial-year reporting work. Skip macros for now.
What to build: take a real sales or transactions file, clean it properly, and produce a one-sheet summary that answers three questions a manager would ask. Keep the file. It becomes the input for month two.
The mistake here is treating Excel as beneath you and rushing it. Analysts who cannot work quickly in Excel get found out in the first week of the job, because the ad hoc request that arrives at 5pm arrives as a spreadsheet.
Month 2: SQL until it is boring
SQL is the single most tested skill in analyst interviews in India, and it stays tested at every level. Give it a month and expect to spend most of that month writing queries rather than reading about them.
The order that works: SELECT and WHERE, then GROUP BY and aggregates, then joins, then subqueries and CTEs, then window functions. Joins are where people quietly go wrong, because a query that silently drops rows still returns a result that looks fine.
What to build: find a database with at least four related tables and write forty queries against it, starting from “how many orders last month” and ending at “rank customers by spend within each city”. Write them; do not copy them.
The checkpoint is unforgiving on purpose. If you cannot write a three-table join with a GROUP BY and a filter without help, month three will not fix it. A structured course such as SQL: Beginner to Advanced at ₹33,110 exists for exactly this month, mostly because having someone review your queries is worth more than another video.
Fees and ratings last checked August 2026; confirm current numbers with the provider before enrolling.
Month 3: one BI tool, one dashboard, one question
Pick Power BI or Tableau and ignore the other. Power BI appears more often in Indian job postings, particularly at service companies and global capability centres, and the concepts transfer if you switch later.
Cover connecting to data, shaping it in Power Query, building a data model with relationships, writing basic DAX measures, and designing a page that a non-analyst can read.
What to build: one dashboard that answers a specific question, such as which product categories are dragging down margin in which cities. Not a “sales dashboard”. A question.
The checkpoint is a person, not a feature list. Show it to somebody who does not work in data and ask them what it says. If they cannot tell you the finding in one sentence, rebuild it. This is the single most common weakness in beginner portfolios: a dashboard covered in charts that makes no argument.
Month 4: Python, and only the parts you need
Now Python, and deliberately not first. Python is where beginners get stuck for months because there is no natural stopping point, so define one: you are learning to clean and analyse data, not to build software.
Cover Python basics, then pandas properly (reading files, filtering, groupby, merge, handling nulls, dates), then matplotlib or seaborn for quick charts. You do not need object-oriented programming, data structures and algorithms, or web frameworks.
What to build: a notebook that takes a genuinely messy CSV, with inconsistent city names and blank fields and dates in three formats, and turns it into something analysable, with your reasoning written between the cells.
If you want this scoped for analysts rather than for developers, Data Analytics and Statistics with Python at ₹38,110 covers this slice and month five together.
Month 5: statistics, and a project you can defend
The statistics an analyst is actually asked about is narrower than most syllabi suggest: descriptive statistics, distributions, correlation against causation, sampling and bias, confidence intervals, and A/B test basics. You do not need calculus or linear algebra for analyst work.
Spend half the month on this and half on a second project that combines everything so far. The project should end in a written conclusion of a few paragraphs, including what you could not conclude and why. Interviewers probe that part hardest, because it is where they find out whether you understand your own analysis.
The mistake to avoid this month is starting machine learning. Predicting churn with a model you cannot explain adds nothing to an analyst application, and it takes time away from the SQL and business reasoning that will actually be tested.
Month 6: portfolio, resume, and volume
Three projects, finished and written up, is the target. Each one needs a one-paragraph problem statement, the data source, what you did, what you found, and what you would do next. Publish them somewhere linkable.
Then apply properly. Fifty applications is a floor, not a stretch, and a referral from someone in the company beats twenty cold applications, so tell the people you already know what you are looking for. Expect the first ten applications to teach you what is wrong with your resume.
Interview preparation runs in parallel: SQL questions daily, one project you can narrate cold, and a plain-English answer to “what would you do if the data looked wrong”. If you have not had a first-round interview by the end of month six, the problem is almost always the resume or the target list rather than your skills.
What this roadmap leaves out on purpose
Machine learning, big data tools like Spark and Hadoop, cloud certifications, R, and advanced statistics. Every one of them appears on somebody’s beginner roadmap and none of them will get a first analyst job in India. Add them later, when a specific role asks for them.
It also leaves out collecting certificates. One structured program with projects and mentorship does more than four self-paced certificates, a case we make with numbers in are data analytics bootcamps worth it in India.
If self-study is not working by month two
Be honest about this early. Self-study fails for most people not on ability but on sequencing and feedback: nobody tells you your SQL is inefficient, your dashboard is unreadable, or that you have been on Python for six weeks too long. If you have restarted twice, a structured live program buys you the sequence and the feedback, and our beginner course comparison sets out what the options cost.
Where a course is the wrong answer: if you want a deep computer science education, or if you learn best entirely at your own pace and have already proved you finish things alone. Live cohorts are built around fixed class timings and deadlines, and that structure is the product.
Frequently asked questions
How long does it take to become a data analyst from scratch?
Six months at 10 to 12 hours a week is realistic for someone working full time, and three to four months if you can give it 20 or more hours a week. The variable is not intelligence, it is consistency and whether you get feedback on your work. People who restart every few weeks routinely take over a year to reach the same point.
What should I learn first as a data analyst?
Excel, then SQL. Excel teaches you the shape of business data and is still where most Indian reporting happens, and SQL is the most tested skill in analyst interviews. Learning Python first is the most common sequencing error, because it takes months to reach competence and does not answer the questions an interviewer asks first.
Do I need Python to get a data analyst job in India?
Not for every role, but it widens the pool considerably, especially at product companies and startups. Many reporting and business analyst roles run entirely on SQL, Excel and Power BI. Treat Python as the month-four investment that opens the better-paying half of the market rather than as a prerequisite for applying.
Is a data analyst roadmap different for non-tech graduates?
The sequence is identical. Commerce and arts graduates usually move faster through Excel and business context, and need an extra week or two on SQL and Python before things click. Since most Indian job postings ask for a graduate in any discipline, the roadmap is the same and only the pacing differs.
Can I follow this roadmap alongside a full-time job?
Yes, and that is what the 10 to 12 hour assumption is built for. The pattern that works is one hour on weekday mornings and a longer block on the weekend for project work, because projects need uninterrupted time in a way that practice questions do not. Protect the project block first when the week gets busy.
How many projects do I need in my portfolio?
Three finished ones beat eight abandoned ones. Each should answer a real question, use data you did not get from a tutorial, and end in a written conclusion. Interviewers spend their time probing one project in depth rather than counting how many you have listed.
Ready to stop restarting
A roadmap solves the sequence problem. It does not solve the feedback problem, which is what actually stops people in month two.
SkilloVilla’s Data Analytics with Python track runs this arc over 4 to 5 months of live classes at ₹71,999, currently ₹58,999, with 1:1 mentorship from working analysts reviewing what you build, real projects and placement support. Alumni have been hired at Accenture, Deloitte, Flipkart, Cred and Razorpay. Book a counselling session and find out where in this roadmap you should actually start.
