Thinking Like an Analyst · Lesson 1 of 6
What Does a Data Analyst Do? The Analytics Workflow Explained
A data analyst turns questions into answers a manager can act on. See a real day of work, the six-step analytics workflow, the four types of analysis and the skills recruiters actually test.
Picture the operations head of a food-delivery company in Hyderabad saying: "Customers are complaining more this month. What is going on?" Nobody hands the analyst a neat dataset and a formula. The job is to turn that worried sentence into a clear answer — "Late deliveries doubled in three areas after we changed the restaurant partner in Kondapur, and 70% of complaints come from those orders" — and a recommendation people can act on.
That is data analytics: using data to answer business questions and support decisions.
A normal week for a fresher analyst
- Monday: refresh the weekly sales dashboard and explain why the North region dropped 8%.
- Tuesday: write SQL to pull last quarter's returns, clean it, and find which products are returned most.
- Wednesday: a manager asks for "a quick number" — active users who ordered twice this month.
- Thursday: build a short slide deck with three findings and one recommendation.
- Friday: check that yesterday's data load did not create duplicate orders.
Notice how little of this is fancy maths. Most of the value comes from asking the right question, getting clean data and explaining the result clearly.
The six-step analytics workflow
- Ask — turn a vague concern into specific questions. "Complaints are up" becomes "Which complaint types increased, in which cities, since when?"
- Collect — find where the data lives: a database, Excel exports, a CRM, survey forms.
- Clean — fix missing values, duplicates, wrong formats. This often takes half the time.
- Analyse — summarise, compare, group, look for patterns and unusual values.
- Share — charts, a dashboard or a short written summary aimed at the person deciding.
- Act — a recommendation, and later a check on whether the action worked.
Remember: a perfect analysis of the wrong question is useless. Spend real time on step 1 — and write the questions down before you open any data.
The four types of analysis
| Type | Question | Example |
|---|---|---|
| Descriptive | What happened? | Sales were ₹42 lakh in October, down 6%. |
| Diagnostic | Why did it happen? | The drop came from two stores that were closed for renovation. |
| Predictive | What is likely to happen? | Diwali week will need 30% more stock. |
| Prescriptive | What should we do? | Move stock from Store A to Store B before the festival. |
Fresher roles are mostly descriptive and diagnostic. Predictive work leads towards data science and machine learning.
Analyst vs data scientist vs data engineer
- Data analyst — answers business questions with SQL, Excel, dashboards and some Python.
- Data scientist — builds predictive models and experiments; more statistics and ML.
- Data engineer — builds the pipelines and databases that deliver clean data to everyone else.
Many people start as analysts and move into either direction after a year or two.
Skills recruiters test
- SQL — almost every analyst interview has a SQL round.
- Excel — pivot tables, lookups and formulas, often in a timed case task.
- A dashboard tool — Power BI, Tableau or Looker Studio.
- Python (pandas) — increasingly expected, especially at product companies.
- Business sense and communication — "So what? What should we do?" is the question you will hear most.
Try it yourself: Take this vague request: "Our college canteen is losing money." Write five specific questions an analyst should ask, and for each one, the data you would need. You have just done the most important step of any analysis.
What to learn next
Real data is messy. The next lesson shows how analysts clean it — and why skipping this step ruins otherwise good work.
Common interview questions
Try answering each one aloud before you open it.
1What does a data analyst do?
Collects, cleans and analyses data to answer business questions, then communicates findings through reports, dashboards and recommendations that help people make decisions.
2Explain the difference between descriptive and diagnostic analytics.
Descriptive analytics summarises what happened, such as monthly sales. Diagnostic analytics explains why it happened, for example by breaking the change down by region, product or customer segment.
3How is a data analyst different from a data scientist?
Analysts focus on answering business questions with existing data using SQL, spreadsheets and dashboards. Data scientists focus more on prediction and experimentation using statistics and machine learning.
Practise this
Build something with it
FAQs
Is data analytics a good career for freshers in India?
Yes. Analyst roles exist in IT services, banks, e-commerce, consulting and startups, and the entry barrier is skills plus a portfolio rather than a specific degree.
Do analysts need to code?
SQL, yes — it is essential. Python is a strong plus. Many analysts work mostly in SQL, Excel and a dashboard tool.
References & further reading
Finished this lesson?
Mark it complete. Create a free account to save it on every device.