Tech

How to Become a Data Analyst From Scratch (Step-by-Step Guide for 2026)

By STEMLaunch Editorial Team · Updated 3 October 2026 · 10 min read

Fact checking: source-led editorial review

Data analyst is one of the most in-demand jobs of 2026, and you don't need a maths degree to get there. Here's exactly how to go from zero to job-ready, step by step.

Data analyst is one of the most searched career-change terms on the internet right now.

And for good reason. It's a role that exists in almost every industry, pays well from the start, has a clear learning path, and, crucially, doesn't require a maths or computer science degree to break into.

What it does require is the right skills, in the right order, with something to show for it. This guide walks you through exactly that, from absolute zero to applying for your first role.

What does a data analyst actually do?

Data analysts collect, clean, and interpret data to help organisations make better decisions. That's the simple version.

In practice, the work varies a lot by industry and company size. But a typical week might include:

• Pulling data from a database using SQL and cleaning it up, removing duplicates, fixing formatting errors, handling missing values
• Building a dashboard in Tableau or Power BI to track key metrics for a team or client
• Writing a report or presentation explaining what the data shows and what the business should do about it
• Answering ad-hoc questions from colleagues: “why did sales drop in Q3?”, “which customer segment is most profitable?”
• Working with engineers or product managers to understand what data is available and how it's collected

It's a role that rewards clear thinking and clear communication as much as technical skill. The people who do it best can translate numbers into plain language that non-technical colleagues actually understand.

It's also worth noting that “data analyst” sits on a spectrum. At the junior end, you're doing a lot of cleaning and reporting. As you progress, you move into more complex analysis, predictive modelling, and eventually into data science if you want to go there.

The skills you need, and the order to learn them

This is the most important part of the guide. A lot of people try to learn everything at once, get overwhelmed, and give up. Don't do that. Learn these in order.

Stage 1: Excel and spreadsheets

Start here. Not because Excel is glamorous, it's not, but because it's used in data roles at every level, and mastering it teaches you the foundational thinking you need before moving to anything more advanced. Learn: VLOOKUP, XLOOKUP, pivot tables, conditional formatting, basic formulas, and data validation. Free resource: ExcelJet.net. Time: 2–4 weeks.

Stage 2: SQL

SQL (Structured Query Language) is the language you use to pull data out of databases. It's the single most important technical skill for a data analyst, and it's more learnable than people expect. Learn: SELECT, FROM, WHERE, GROUP BY, JOIN, ORDER BY, subqueries. Free resources: SQLZoo, Mode Analytics SQL Tutorial, Khan Academy. Time: 4–8 weeks to get job-relevant.

Stage 3: Data visualisation

Once you can pull and manipulate data, you need to present it. Learn one visualisation tool properly before touching others. Tableau Public is free and widely used. Power BI is free and dominant in corporate environments. Pick one, build dashboards with real datasets, and publish them publicly. Time: 3–5 weeks to get comfortable.

Stage 4: Python or R (pick one)

You don't need to become a programmer. But knowing enough Python to clean data with Pandas, run basic statistical analysis with NumPy, and create charts with Matplotlib puts you ahead of a significant chunk of junior applicants. Python is the better choice for most people, it's more versatile and more widely used outside analytics. Free resources: freeCodeCamp Python for Data Analysis, Kaggle's Python and Pandas courses. Time: 6–10 weeks to useful proficiency.

Stage 5: Statistics basics

You don't need a statistics degree. You need to understand mean, median, mode, standard deviation, correlation vs causation, and basic probability. These come up constantly in real analysis work and in interviews. Free resource: Khan Academy Statistics. Time: 2–3 weeks.

Total timeline if you study consistently (1–2 hours a day): 4–8 months to job-ready.

The best resources in 2026

How to build a portfolio

A portfolio is what separates candidates who get interviews from candidates who don't. Here's how to build one without any job experience.

Use public datasets.

Kaggle, Data.gov (US), data.gov.uk, the World Bank, and Our World in Data all have free, real datasets on topics from climate to crime to sport. Pick something you're genuinely interested in, the curiosity shows in the quality of the analysis.

Do the analysis end-to-end.

For each project: download the data, clean it, analyse it, visualise it, and write up what you found and what you'd recommend. That full cycle is what employers want to see. A half-finished notebook with no conclusions isn't a portfolio piece.

Publish everything.

GitHub for your code and notebooks. Tableau Public for your dashboards. A personal website or Notion page linking to both. Everything public, everything linkable. Put the links in your CV.

Aim for three strong projects.

Three well-documented, interesting projects beat ten half-finished ones every time. Quality over quantity. Each one should answer a real question with real data.

Kaggle competitions.

Even finishing in the bottom half of a Kaggle competition and writing up your approach shows you can work with messy real-world data and think analytically. Enter one, document your process, and include it in your portfolio.

Realistic salary expectations

How to get your first role

Once your portfolio has three solid projects and you're comfortable with SQL, Excel, and one visualisation tool, you're ready to start applying. Here's how to approach it:

Target the right job titles.

Junior data analyst, data analyst, business analyst, reporting analyst, and insights analyst are all worth searching. Business analyst roles often have the lowest technical bar and are a good entry point if you're struggling to land a pure data role first.

Tailor your CV for each application.

Look at the job description, identify the tools and skills they mention (SQL, Tableau, Power BI, Python), and make sure those appear prominently in your CV and cover letter. One tailored application beats ten generic ones.

Use LinkedIn properly.

Connect with data analysts in companies you're interested in. Comment thoughtfully on posts. Share your portfolio projects as LinkedIn posts with a short write-up of what you found. This builds visibility organically.

Apply to apprenticeship programmes.

Data apprenticeships exist at major employers including PwC, Capgemini, the NHS, and various government departments. They're paid, structured, and specifically designed for people without prior experience. Look for them on apprenticeship portals in your country.

Don't wait until you feel ready.

Most junior data job descriptions list ten requirements. Employers don't expect candidates to meet all of them. If you meet six or seven, apply. The worst they can say is no, and the process of applying and interviewing teaches you things you can't learn from courses.

The bottom line

Data analyst is one of the most achievable career pivots available in 2026. The demand is real, the tools are learnable, and the path from zero to employed, while not instant, is clear and well-trodden.

Most people who successfully make this transition do so within 6–12 months of consistent effort. The ones who don't either spread themselves too thin across too many tools, or wait until they feel “ready” before building a portfolio.

Start with Excel. Then SQL. Build something. Put it on GitHub. Apply before you think you're ready.

About the STEMLaunch Editorial Team

STEMLaunch is an independent UK careers education publication. Our editorial team researches career routes, qualifications and pay using official and primary sources.

LinkedInShare on XEmail