How to learn data analytics from scratch works best in stages: master Excel and basic statistics first, move on to SQL for pulling data, then visualisation with Power BI or Tableau. Python can follow as a complement. Each stage is practised on real datasets, so your ability grows from doing.
- Begin with Excel and basic statistics before touching more complex tools
- SQL is the most requested skill in data analyst job postings
- One real project teaches more than ten passive tutorial videos
- A laptop with Microsoft Excel or Google Sheets
- Internet access for online SQL practice and downloading datasets
- One public dataset that genuinely interests you
- A steady study schedule, for example five to ten hours per week
The Demand for Data Analytics in Numbers
What a data analyst actually does
A data analyst turns raw data into decisions. They pull data from company databases, clean the messy parts, look for patterns, and present findings in a form managers can grasp quickly. The work rests on curiosity and care. Advanced mathematics rarely sits at its core. Many beginners picture data analytics as full of complex algorithms. The daily reality is simpler and more valuable: making sure sales figures are correct, finding why a campaign underperformed, or building a weekly report people can trust. The skill of translating numbers into a clear story is exactly what workplaces value most. That is why the learning path starts with the tools used most often in practice. Excel to understand data structure, SQL to pull it from databases, and visualisation tools to communicate results. This foundation applies across industries, from banking to digital startups in Indonesia.
7 Steps to Learn Data Analytics from Scratch
These seven steps map the journey from a beginner with no technical background to an aspiring data analyst with a portfolio to show. Work through them in order and practise each stage with real data.
- 1
Master Excel or Google Sheets until fluent
Start with the most accessible tool. Excel and Google Sheets teach you to think about data in rows and columns, the basis of every tool that follows. Learn filtering, sorting, formulas such as VLOOKUP and INDEX MATCH, pivot tables for summarising data, and basic charts. Pivot tables in particular train you to group and condense thousands of rows into meaningful summaries. A good exercise is taking your own spending data or a simple sales record and answering a concrete question from it. Fluency in Excel builds the confidence you need before moving to more technical tools.
Tips- Master pivot tables first, this skill is used most often at work
- Practise each new function directly on data you already understand
- 2
Build a foundation in descriptive statistics
The statistics a beginner data analyst needs are basic and learnable by anyone. Understand mean, median, and mode along with when each fits. Get to know spread through range and standard deviation, and how to read percentages and proportions without slipping up. The idea of correlation helps you see relationships between variables, while reminding you that a relationship does not always mean cause and effect. This foundation keeps your analysis honest and prevents misleading conclusions.
Tips- Learn statistics through real examples so the concepts stick
- Always ask whether an average fairly represents the whole dataset
Be careful about inferring cause and effect from correlation alone. Two figures moving together do not necessarily influence each other. - 3
Learn SQL to retrieve data
SQL is the language for talking to databases, and it is the most requested skill in data analyst postings. At work, data rarely arrives in tidy files. It lives in large databases, and SQL is how you pull the part you need. Learn the core commands SELECT, WHERE, GROUP BY, ORDER BY, and JOIN to combine multiple tables. Aggregate functions such as COUNT, SUM, and AVG round out your ability to summarise data straight from the source. Practise on free training databases so writing correct, efficient queries becomes second nature.
Tips- Practise JOIN seriously, many interview questions centre on it
- Use free online practice databases to sharpen queries every day
- 4
Learn visualisation with Power BI or Tableau
Good analysis is wasted if the result is hard to grasp. Visualisation tools such as Power BI and Tableau turn tables of numbers into dashboards that tell a story. Learn to choose the right chart type: bar charts for comparison, lines for trends over time, and maps for regional data. Build dashboards that answer one business question cleanly without clutter. Tableau Public and the free version of Power BI are enough to practise and build early work. The skill of communicating findings visually often sets candidates apart when applying for a first role.
Tips- One dashboard should focus on answering a single main question
- Avoid too many colours and elements that blur the message
- 5
Add Python as a complement
Once the three foundations above are solid, Python widens your reach. For a data analyst, it is enough to learn the pandas library for handling data, Matplotlib or Seaborn for charts, and the basics of reading and filtering data. Python helps when data cleaning becomes too repetitive to do by hand, or when you want to automate a routine report. This stage is optional early in your career. Many productive data analysts work mainly with SQL and Excel, then add Python as needs grow.
Tips- Focus on pandas for analysis, postpone advanced programming topics
- Learn Python through real data projects so it feels relevant
- 6
Complete a real analysis project end to end
Theory settles into skill when you work through a full project. Pick a public dataset that interests you, for example city transport data, commodity prices, or online sales trends. Frame one question you want to answer, pull the data with SQL or download it directly, clean it in Excel or Python, then present findings in a dashboard. One project seen through to the end teaches the real workflow, from messy data to a defensible conclusion. This experience sticks far more than watching tutorials passively.
Tips- Pick a topic you genuinely care about so you finish it
- Note every analysis decision so it is easy to explain in interviews
- 7
Build a portfolio and prepare to apply
Gather your three or four best projects into a portfolio anyone can access. Upload notebooks and files to GitHub, publish dashboards on Tableau Public, and write a short summary for each project: the question, data source, method, and key finding. A well-arranged portfolio proves your ability far more convincingly than a list of certificates. While building the portfolio, practise explaining projects out loud, since data analytics interviews often test the thinking behind your numbers.
Tips- Link your portfolio at the top of your CV so recruiters find it fast
- Prepare a one-minute story for each standout project
Four Core Tools Beginners Should Master
Excel or Google Sheets
FoundationThe friendliest starting point for understanding data structure, pivot tables, and basic formulas. Still used daily in many data roles.
SQL
CoreThe language for pulling and summarising data from databases. The skill most frequently requested in data analyst postings.
Power BI or Tableau
VisualisationVisualisation tools that turn numbers into interactive dashboards. They help you present findings to decision makers.
Python
AdvancedA complement for automation and larger-scale data handling through the pandas library. Useful once other foundations are in place.
Data Analyst, Data Scientist, and Business Intelligence
| Aspect | Data Analyst | Data Scientist |
|---|---|---|
| Main focus | Explaining what happened from historical data | Predicting and modelling future events |
| Typical tools | SQL, Excel, Power BI, Tableau | Python, machine learning, advanced statistics |
| Learning entry point | Beginner friendly, no technical background needed | Requires deeper programming and mathematics foundations |
| Typical output | Reports, dashboards, business recommendations | Predictive models, algorithms, experiments |
Business intelligence sits close to the data analyst role, with a stronger emphasis on dashboards and routine reporting. Many people begin as data analysts, then grow according to their interests.
“The beginners who progress fastest are usually the ones who get their hands dirty with real data in the first week. They learn SQL while answering genuine questions, so each command gains immediate context. The flow is simple: pull the data, clean it, find the pattern, tell the story.”
Readiness Checklist for Your First Data Role
- You can filter, sort, and summarise data with Excel pivot tables fluently
- You can write SQL queries with JOINs and aggregate functions unaided
- You can build a dashboard that answers one business question in Power BI or Tableau
- You have three or four analysis projects that are publicly accessible
- You can explain the thinking behind each project concisely and clearly
Self-Study or Structured Guidance
- A structured order of topics so you are never unsure where to start
- A mentor who corrects your queries and analysis directly
- Feedback on projects that speeds up how your thinking improves
- A kept schedule so the learning process does not easily stall
- Easy to get lost choosing material among thousands of free sources
- Analysis mistakes often go unnoticed with no one to correct them
- Motivation fades quickly when a hard concept is faced alone
- Time lost on topics that are not yet relevant at the early stage
- Learning data analytics from scratch works best in the order Excel, basic statistics, SQL, then visualisation.
- SQL is the skill recruiters request most, appearing in more than half of data analyst job postings.
- A portfolio of three or four real projects proves ability far more convincingly than certificates.