Data Analyst Roadmap: How to Become an Analyst

The plan in five steps

A data analyst turns data into answers to business questions: why sales dropped, which customers leave, which campaign worked. Learn the tools in this order:

1. Excel or Google Sheets.
2. SQL.
3. Statistics.
4. Visualization and BI.
5. Python for analysis.

Each step extends the previous one: from a single spreadsheet to databases, conclusions and automation.

Step 1. Spreadsheets

• Formulas, relative and absolute references.
• Lookup functions (XLOOKUP or VLOOKUP), IF, SUMIFS, COUNTIFS.
• Pivot tables and filters.
• Data cleaning: duplicates, missing values, inconsistent date formats.

Checkpoint: you can take a raw sales export and quickly build a summary by month, category and region.

Step 2. SQL

SQL is an analyst's main everyday tool.

• SELECT, WHERE, ORDER BY, aggregate functions, GROUP BY and HAVING.
• Different JOIN types and knowing when rows get duplicated.
• Subqueries and CTEs.
• Window functions: ranking, running totals, period-over-period comparison.

WITH monthly AS (
  SELECT DATE_TRUNC('month', order_date) AS month,
         SUM(amount) AS revenue
  FROM orders
  GROUP BY 1
)
SELECT month,
       revenue,
       revenue - LAG(revenue) OVER (ORDER BY month) AS diff
FROM monthly
ORDER BY month;

Checkpoint: you confidently solve join and window-function tasks and sanity-check your results.

Step 3. Statistics

• Mean, median, spread, outliers.
• Distributions and sampling.
• Correlation vs causation: why one does not prove the other.
• A/B testing basics: hypothesis, statistical significance, sample size.

Checkpoint: you can explain why an average order value can mislead and when the median is a better choice.

Step 4. Visualization

• Matching the chart to the question: trend, comparison, share, distribution.
• BI tools: Power BI, Tableau, Looker Studio or similar.
• Dashboards: a few key metrics instead of dozens of charts.
• Storytelling: the conclusion first, then the evidence, then the recommendation.

Checkpoint: someone without a technical background grasps the main takeaway of your dashboard within a minute.

Step 5. Python

• Language basics and Jupyter Notebook.
• pandas to load, clean and group data.
• matplotlib or seaborn for charts.
• Automating recurring reports.

import pandas as pd

df = pd.read_csv("orders.csv", parse_dates=["order_date"])
monthly = df.groupby(df["order_date"].dt.to_period("M"))["amount"].sum()
print(monthly.tail())

Portfolio case ideas

Use public data: government open data portals, Kaggle, public city transport or weather datasets.

For each case, write a short report: question, data, method, conclusion, limitations, recommendation.

Common mistakes

• Learning Python before SQL and stumbling on simple queries.
• Building pretty charts that answer no question.
• Drawing causal conclusions from a single correlation.
• Not checking data for missing values and duplicates.