Data Analyst Skills Employers Ask For: SQL, Excel, BI and More

Open ten data analyst job posts and put the "required skills" lines side by side. The same short list appears again and again: SQL, Excel, a dashboard tool, some statistics, and the ability to explain a result to a person who does not like numbers. The rest is decoration.

This post covers those skills only: what each one means on the job, how an interviewer checks it, and what to learn first. It does not cover the full career path. For the steps from 12th standard to a first job, see the data analyst roadmap after 12th commerce.

Employers ask data analysts for SQL, Excel, a visualization tool such as Power BI or Tableau, basic statistics, and clear communication. Python or R is a common extra. Of these, SQL and Excel come up in almost every posting, so start there.

Skills at a glance

SQL: the skill that gets you to the data

Most company data sits in databases, and SQL is how you ask for it. A task like "revenue by city, highest first" is the kind of question you are expected to answer in one query.

SELECT city,
       COUNT(*) AS orders,
       ROUND(SUM(amount), 2) AS revenue
FROM orders
WHERE amount IS NOT NULL
GROUP BY city
ORDER BY revenue DESC;

I ran this on a small sample table in SQLite. It returned Pune with 3 orders and 1500.0 revenue, then Delhi with 1 order and 800.0. Notice the IS NOT NULL line. Missing values are the reason many beginner queries give wrong totals.

Interviewers usually test joins, GROUP BY, HAVING, subqueries and window functions. Start with the free tutorial in learn SQL for beginners, then practise questions from the SQL window functions interview guide.

Excel: still the fastest tool for small data

Excel is where analysts check a file in two minutes before writing any code. The usual test is a set of formulas on a messy sheet: XLOOKUP or VLOOKUP, SUMIFS, COUNTIFS, IF, text clean-up and pivot tables.

1. Clean

   Remove blanks and duplicates, fix dates stored as text.

2. Summarise

   Build a pivot table by month and category.

3. Chart

   Add one chart that answers one question.

Pick a free sample file, such as a sales sheet from your own college fest, and do those three steps. Our Excel interview questions for data analysts give you a practice set of formulas.

Power BI or Tableau: turning numbers into a dashboard

A dashboard is a page of charts that updates when the data changes. Employers want to see that you can connect a data source, build a few clear visuals and add filters. Learn one tool well. Microsoft's Power BI documentation is free and has guided paths. Tableau also offers free training videos.

Keep each dashboard simple. One page, one question, a title that states the answer, such as "February sales rose because of Pune". A page of twelve charts with no message shows the reader nothing.

Statistics: knowing when a number can be trusted

You do not need advanced maths. You do need the basic ideas well enough to explain them.

• Mean and median. Use the median when a few very large values pull the average up.
• Spread. Standard deviation tells you how far values sit from the average.
• Sampling. A result from 20 customers may not describe all customers.
• Correlation. Two numbers rising together does not mean one causes the other.
• A/B test basics. Know what a p-value is trying to tell you.

An interviewer may ask: "Sales went up after the new banner. Did the banner cause it?" A good answer asks what else changed that week and whether a control group exists.

Python or R: for data that is too big for Excel

Python is the more common choice in India and the US. The pandas library handles tables with millions of rows and lets you repeat a clean-up with one command instead of clicking again. Here is a short example that fills missing amounts with the median and totals sales by month.

import pandas as pd
df = pd.read_csv("orders.csv")
df["amount"] = df["amount"].fillna(df["amount"].median())
df["order_date"] = pd.to_datetime(df["order_date"])
monthly = df.groupby(df["order_date"].dt.to_period("M"))["amount"].sum()
print(monthly)

I tested the same steps on the five-row sample table above. January came out at 1300 and February at 1600. Fill rules like this hide choices: filling with the median changes the totals. Write down what you chose and why.

Data cleaning: the unglamorous skill

Real data arrives with blanks, duplicates, wrong dates and city names spelt three ways. Most of an analyst's time goes into fixing these before any chart is possible, so employers ask about it directly. A strong answer lists the checks you run every time: count rows, count blanks per column, look for duplicates, check the minimum and maximum of every number.

Communication and business sense

This is the skill that separates a person who produces tables from one who gets asked to the meeting. Say what you found in one sentence. Say what you think should happen next. Then show the chart that proves it.

Business sense means knowing what the company earns money from. A retail analyst who understands returns and discounts will ask better questions than one who only knows formulas. Our business analyst skills guide covers the overlap with that role.

How to show these skills on a resume

List tools in a skills line, then prove each one in a project. A project that shows all five skills is a dashboard built from a public dataset, with a short write-up. Pull the data with SQL, clean it in Excel or Python, chart it in Power BI, and state two findings in plain English. Publish the files and the write-up on GitHub, using the steps in the GitHub README guide.

Common mistakes

• Learning five tools shallowly. Depth in SQL and one dashboard tool beats a long list.
• Skipping missing values. A blank can change a total without any error message.
• Charts without a point. Each chart should answer one question that you can write down.
• Starting with machine learning. Entry-level analyst work rarely needs it. Learn it after the basics.

What to do next

Choose a public dataset today and write three questions you want it to answer. Answer the first one in SQL, the second in Excel and the third in a dashboard. You will find your weakest skill within a weekend, and that is the one to study next.