Four skills get you shortlisted for a data analyst job in India: SQL, Excel, one BI tool (Power BI or Tableau), and enough statistics to know when a number is meaningless. Python is a strong fifth and is becoming standard, but plenty of people still get hired without it.
What most skill lists get wrong is the level. “Know SQL” is useless advice. This article says what depth of each tool an Indian employer actually tests, and what you can safely ignore.
The shortlist, with the level expected
This table is the honest version of a job ad’s requirements section. The level column is what an interview will actually probe, not what the ad claims.
| Skill | Level expected for a first job | How it is tested | Priority |
|---|---|---|---|
| SQL | Intermediate: joins, aggregation, subqueries, window functions | A live query round on a sample schema | Non-negotiable |
| Excel | Advanced: pivot tables, lookups, conditional logic, Power Query | A file to clean and summarise, sometimes timed | Non-negotiable |
| Power BI or Tableau | Working: build one end to end from raw data | A portfolio dashboard you have to walk through | Non-negotiable |
| Statistics | Descriptive stats, distributions, correlation, basic significance | Verbal questions on a case, rarely a formal test | High |
| Python | Basic pandas: read, clean, group, merge, plot | A take-home task or a code walkthrough | Rising fast |
| Business communication | Explain a finding to someone who does not care how you got it | The final round, almost always | Underrated |
| Machine learning | None required for most analyst roles | Occasionally asked as a differentiator | Optional |
Employers rarely care whether you use Python or R, as long as the analysis is right and you can defend it. They care a great deal about SQL, because it is the one skill that cannot be faked in a live round.
SQL: where most candidates fall short
SQL was designed at IBM in the early 1970s and is still the single most valuable thing on an analyst’s CV. Almost every Indian data team keeps its data in a warehouse you query with SQL, so this is the skill you use on day one and every day after.
The bar for a first job is not “SELECT and WHERE”. It is the level below, which most candidates who claim intermediate SQL cannot actually reach under interview pressure.
| Skill | What it does | Syntax | Example |
|---|---|---|---|
| Aggregation with grouping | Summarises rows into one line per category | SELECT col, SUM(x) FROM t GROUP BY col | SELECT city, SUM(amount) FROM orders GROUP BY city |
| INNER JOIN | Keeps only rows matching in both tables | SELECT … FROM a JOIN b ON a.id = b.id | SELECT c.name, o.amount FROM customers c JOIN orders o ON c.id = o.customer_id |
| LEFT JOIN | Keeps all rows from the first table | SELECT … FROM a LEFT JOIN b ON a.id = b.id | SELECT c.name, o.amount FROM customers c LEFT JOIN orders o ON c.id = o.customer_id |
| HAVING | Filters after grouping, unlike WHERE | SELECT col, COUNT(x) FROM t GROUP BY col HAVING COUNT(x) > n | SELECT city, COUNT(id) FROM orders GROUP BY city HAVING COUNT(id) > 100 |
| CTE | Names a subquery so the main query stays readable | WITH name AS (SELECT …) SELECT … FROM name | WITH big AS (SELECT customer_id FROM orders WHERE amount > 50000) SELECT COUNT(1) FROM big |
| Window function | Ranks or runs totals without collapsing rows | SELECT col, RANK() OVER (PARTITION BY a ORDER BY b DESC) FROM t | SELECT city, amount, RANK() OVER (PARTITION BY city ORDER BY amount DESC) FROM orders |
If you can write all six of those without looking anything up, you are past the SQL round at most Indian analytics employers. Window functions are the usual failure point, and they are also the ones that come up in every second interview.
Excel: still the first tool you are tested on
People underestimate Excel because it feels like a school subject. Indian hiring managers do not. In services firms, consulting and BFSI, a large share of real analysis still happens in a spreadsheet, and an analyst who is slow in Excel is visibly slow. The functions below are the ones that turn up in actual analyst tests.
| Function | What it does | Syntax | Example |
|---|---|---|---|
| SUMIFS | Adds values that meet several conditions | =SUMIFS(sum_range, range1, criteria1, range2, criteria2) | =SUMIFS(C2:C500, A2:A500, “Bengaluru”, B2:B500, “Aug”) |
| COUNTIFS | Counts rows meeting several conditions | =COUNTIFS(range1, criteria1, range2, criteria2) | =COUNTIFS(A2:A500, “Pune”, D2:D500, “>10000”) |
| XLOOKUP | Finds a value in one column and returns another | =XLOOKUP(lookup_value, lookup_array, return_array) | =XLOOKUP(F2, A2:A500, C2:C500) |
| INDEX with MATCH | The older lookup that still appears in tests | =INDEX(return_range, MATCH(value, lookup_range, 0)) | =INDEX(C2:C500, MATCH(F2, A2:A500, 0)) |
| IFERROR | Replaces an error with something readable | =IFERROR(expression, value_if_error) | =IFERROR(XLOOKUP(F2, A2:A500, C2:C500), “Not found”) |
| TEXT | Reformats a date or number for a report | =TEXT(value, format_text) | =TEXT(A2, “dd-mmm-yyyy”) |
Beyond formulas, learn pivot tables properly and learn Power Query. Power Query is the part most self-taught analysts skip, and it is what turns a two-hour monthly cleaning job into a refresh button.
Power BI or Tableau: pick one and go deep
Do not learn both. Indian job ads name one or the other, Power BI more often because of Microsoft licensing across enterprise India, and a hiring manager would rather see one dashboard you built end to end than screenshots from two tools.
What “working level” means: connect to a source, shape the data before it reaches the visual layer, build a model with relationships, write measures (DAX in Power BI, calculated fields in Tableau), and design a page someone can read in ten seconds. That last part is where most portfolio dashboards fail. Twelve charts on one page is not a dashboard.
Python: when it starts to matter
You can get a first analyst job in India without Python, particularly in BI-heavy and MIS-adjacent roles. You will hit a ceiling within two years without it.
The level worth reaching before you apply is narrow and achievable: pandas for reading files and databases, cleaning columns, grouping, merging DataFrames, and handling dates; plus matplotlib or seaborn to produce a chart. You do not need object-oriented programming, algorithms or web frameworks. If a tutorial starts teaching you classes and decorators, you have wandered into a software engineering course.
Our Data Analytics & Statistics with Python course at ₹38,110 covers exactly this scope, and the SQL: Beginner to Advanced course at ₹33,110 covers the query skills above.
Statistics: how much you really need
Enough to avoid embarrassing yourself, which is less than a statistics degree and more than nothing. Specifically: mean versus median and when the mean lies, standard deviation, what a distribution looks like when it is skewed, correlation and why it is not causation, sampling, and the idea behind a significance test.
You need this less for calculations and more for judgement. The most common analyst mistake in India is not a wrong formula, it is presenting an average from 40 rows as a finding, or reporting a 12 percent lift that sits inside the noise.
The half nobody trains for
Critical thinking. The business rarely hands you the right question. They ask why revenue dropped; the useful answer usually requires reframing that into three narrower questions you can actually test.
Communication. You will explain a result to a category manager who has no interest in your method. If you cannot get to the point in two sentences, the analysis does not travel.
Presentation. Not slide aesthetics, but sequencing: what the number is, why it moved, what you recommend, in that order. This is learnable and most analysts learn it badly by trial and error.
Stakeholder scepticism. Being asked “are you sure?” in a room of senior people is routine, and knowing your own data well enough to answer calmly is a skill.
What you can safely skip
Machine learning is not a data analyst skill in India for a first job. It is a differentiator at senior analyst level and a requirement for data scientists, and learning it early usually costs you the SQL depth that would actually get you hired.
Linear algebra and calculus are not prerequisites for a data analyst, whatever older career articles claim. They matter for data science and for understanding how algorithms work. If someone tells you to do a maths postgraduate degree before applying for an analyst role, they are describing a different job.
Big data tooling such as Spark, and cloud certifications, can also wait. So can R, unless a specific employer asks for it.
What the ladder looks like
| Level | SQL | Excel | BI tool | Python | Extra |
|---|---|---|---|---|---|
| Junior analyst | Joins, aggregation, subqueries | Pivots, lookups, conditional logic | Can build a basic dashboard | Optional | Clean documentation habits |
| Analyst, 2 to 4 years | Window functions, query tuning | Power Query, complex models | Data modelling and measures | pandas fluency | Owns a metric definition |
| Senior analyst | Warehouse-level design awareness | Used sparingly by now | Advanced DAX, row-level security | Automation and scripting | Experiment design, stakeholder management |
The pattern is worth noticing: Excel matters most at the start and least later, while SQL depth keeps compounding for your whole career. Spend your learning time accordingly.
Frequently asked questions
What are the most important skills for a data analyst in 2026?
SQL, Excel, one business intelligence tool such as Power BI or Tableau, and working statistical judgement. Python is close behind and is now expected in most product-company and analytics-firm roles. Communication is the skill that separates analysts who get promoted from those who stay at the same level.
Can I become a data analyst without coding?
Partly. SQL is a query language and is much simpler than a general-purpose programming language, but you cannot skip it. What you can skip at the start is Python, particularly for BI-heavy roles built around Power BI or Tableau, though you will want it within a couple of years.
Do data analysts need machine learning?
No, not for a first job in India. Machine learning is a data science requirement and a senior-analyst differentiator, and beginners consistently overinvest in it at the cost of SQL depth. Learn it once you are employed and can see whether your team actually uses it.
Is Excel still relevant for data analysts?
Very much so, especially in Indian services firms, consulting and financial services, where a lot of real analysis still happens in spreadsheets. Interviews commonly include a timed Excel exercise on a messy file. Advanced Excel plus Power Query is a faster route to a first job than an extra programming language.
Should I learn Power BI or Tableau?
Learn one, and let the job ads in your target city decide which. Power BI appears more often in Indian enterprise roles because of existing Microsoft licensing, while Tableau shows up more in product companies and some analytics firms. The underlying skills transfer, so the second tool takes weeks rather than months once you know the first.
How long does it take to learn data analyst skills?
Four to six months of consistent daily effort is realistic for someone starting from zero, covering SQL, Excel, a BI tool and basic Python alongside two or three portfolio projects. People who try to compress it into six weeks usually have the tool syntax but no ability to handle a messy dataset or defend a finding, which is what interviews test.
Learn these in the order that gets you hired
The sequence matters as much as the list. SQL first, Excel alongside it, then a BI tool, then Python, with statistics threaded through all of it and at least two projects on real messy data.
SkilloVilla’s Data Analytics with Python track is ₹71,999, currently ₹58,999, and runs that sequence over four to five months of live classes, with 1:1 mentorship from working analysts, real projects and placement support. Hiring partners include Accenture, PwC, Deloitte, Flipkart, Swiggy and Razorpay. Alumni report packages between 4 and 15 LPA, with a median of 9.5 LPA. Scholarships are available depending on profile. If you want the entry requirements rather than the skill list, our guide to data analyst qualifications in India covers what employers check on a CV.
Fees and ratings last checked August 2026; confirm current numbers with the provider before enrolling.
