If you want to know how to become a data analyst, start with one fact: there is no licence, compulsory exam or single required degree for the job. Employers hire analysts because they can show they will take a messy table, ask the right question of it and explain the answer to someone who has to make a decision. Your task is to build proof of that, and the fastest route depends on where you are starting.
This guide is organised around three starting points. Find yours, then use the shared sections on skills, portfolio and job hunting.
What a data analyst actually does all week
Job titles vary (reporting analyst, business intelligence analyst, operations analyst, insights analyst), but the work follows a similar loop:
- Clarify the question. “Why are sales down?” becomes “Which product lines and regions fell in Q3 compared with Q2, and did the price change in August coincide with the drop?”
- Find and extract the data. Usually from a database with SQL, an export from a CRM or finance system, or a spreadsheet someone has been maintaining by hand.
- Clean it. Duplicates, blank fields, inconsistent date formats, product codes that changed halfway through the year. This is often most of the job.
- Analyse. Summaries, comparisons, trends, sometimes a simple statistical test.
- Present. A dashboard, a short written note or a slide that leads with the answer.
The US Bureau of Labor Statistics (BLS) does not publish a separate occupation called “data analyst”. Its closest profile, data scientists, describes very similar duties: identifying useful data, cleaning and analysing it, and recommending business decisions to stakeholders, using tools such as data visualisation software and data-oriented programming languages. Data scientist roles typically go further into machine learning models; many analysts move in that direction later.
Data analyst qualifications: what employers really ask for
Look at ten adverts for junior analyst roles and you will usually see three layers:
- A degree. Many adverts ask for one, often in a quantitative subject (maths, statistics, economics, computer science, engineering) or in business. The BLS says data scientists typically need at least a bachelor’s degree in mathematics, statistics, computer science or a related field. Analyst roles, especially in operations and reporting, are often more flexible, but a degree still widens the pool of jobs you can apply for.
- Tools. Excel or Google Sheets, SQL and one visualisation tool (Power BI and Tableau are common choices). Python or R appears more often as you move up.
- Evidence. Previous analytical work, a portfolio, or a certificate that shows structured learning.
A certificate does not replace experience, but it can be a useful structure for self-study and a signal on a CV. Two widely used options, as described on their providers’ own pages (checked November 2026):
| Credential | What it covers | Format | Good for |
|---|---|---|---|
| Google Data Analytics Certificate (on Coursera) | Spreadsheets, SQL, Tableau, R, presentation tools; eight courses | Self-paced; Google says most learners finish in three to six months at under ten hours a week, with no prior experience needed | Complete beginners who want a guided curriculum |
| Microsoft Certified: Power BI Data Analyst Associate (exam PL-300) | Preparing, modelling, visualising and analysing data, and managing and securing Power BI | Proctored exam, 100 minutes; renewed free every 12 months through an online assessment | People targeting employers that use Microsoft tools, or who already know the basics |
Prices change and vary by country, so check the provider’s page for your location before you enrol.
Find your starting point
Starting point A: you already work with numbers in another job
Finance assistants, sales coordinators, supply-chain staff, HR administrators and marketing executives often do analysis without the title. This is the strongest starting position because you have domain knowledge and access to real problems.
- Volunteer for the reporting nobody wants: the weekly sales pack, the absence tracker, the stock reconciliation.
- Rebuild one manual spreadsheet so it refreshes from a data source rather than being retyped. Record the hours saved.
- Learn SQL so you can query the system yourself rather than waiting for an export.
- Ask to sit in on the analytics or BI team’s work, then apply internally. Internal moves avoid the “no experience” filter.
Starting point B: you are a student or recent graduate
Your degree subject matters less than what you did with data during it. A dissertation with a real dataset, a statistics module, a society you ran with a membership spreadsheet: all of it counts if you describe it in analyst terms.
- Pick two or three modules or projects and rewrite them as analysis stories: question, data, method, finding.
- Apply for graduate schemes and internships in finance, consulting, retail, logistics and government, not only “data” companies.
- Add one tool you have not used in your course, usually SQL or Power BI.
Starting point C: you are starting from scratch
No degree in a related subject and no analytical job? The route is longer, but it exists. Expect to spend several months learning before applying, and to start in a role that is analyst-adjacent: reporting assistant, data entry with reporting duties, junior MI (management information) analyst, operations coordinator.
- Follow a structured course rather than random tutorials, so you learn in a sensible order.
- Build the portfolio below as you learn; do not leave it until the end.
- Use your current job, however unrelated, as a data source. A retail worker can analyse footfall or waste; a driver can analyse route times.
Data analyst skills, in the order to learn them
Trying to learn Python, statistics, SQL and three dashboard tools at once is a common reason beginners stall. This order builds each skill on the last:
- Spreadsheets properly. Lookups, pivot tables, conditional logic, clean table structure, basic charts. Most analysts still use spreadsheets every day.
- SQL. Selecting, filtering, joining tables, grouping and aggregating. This is the skill that most clearly separates analysts from general office roles.
- One visualisation tool. Power BI or Tableau. Choose by checking which appears more in adverts where you want to work.
- Basic statistics. Averages and medians, distributions, correlation versus causation, sampling, and what a significant difference means.
- Python or R. Useful for repeatable cleaning and larger datasets. Learn it once you are comfortable with the first four.
Alongside the tools, two skills decide whether people trust your work: checking (do the totals reconcile with the source system?) and writing (can a manager understand your finding in three sentences?).
Build proof: a three-project portfolio
A portfolio replaces the experience you do not yet have. Three well-explained projects beat ten notebooks with no commentary. The briefs below are illustrative; adapt them to an industry you want to work in.
| Project | Skills it shows | What to publish |
|---|---|---|
| Clean and reconcile a messy public dataset (for example, open government data on transport or trade) | Spreadsheets or Python cleaning, documenting assumptions | Before-and-after sample, a list of every cleaning decision and why |
| Answer a business question with SQL across several tables | Joins, aggregation, structuring a query | The queries, the result table and a 200-word finding |
| Build a one-page dashboard for a named audience (for example, a store manager) | Visualisation, choosing what to leave out | Screenshots, the questions the dashboard answers, one decision it supports |
For each project, write a short summary that leads with the finding, not the method. “Weekend orders fell 18% after the delivery fee change” (illustrative) is a better opening than “I used a LEFT JOIN”.
Getting the first data analyst role
Read adverts for signals
Junior adverts that mention a team, a named reporting tool and specific business questions usually mean you will be trained. Adverts that list every tool in existence for a junior salary usually mean the employer has not decided what it wants. Apply to the first kind first.
Translate your CV into analyst language
Each bullet should show a question, the data, the tool and the outcome. “Built a weekly stock report in Excel from three supplier files, cutting preparation time from four hours to 30 minutes” (illustrative) works; “Responsible for reports” does not. Browse resume examples for analytical roles for ways to phrase this, and if you want a layout already built around analyst skills, the data analyst resume builder gives you sections for tools, projects and results. Before sending, run the file through an ATS resume score checker to make sure tool names such as SQL and Power BI are being read.
Prepare for practical tests
Many analyst interviews include a take-home task or a live exercise in Excel or SQL. Practise timed exercises, and be ready to explain your checking steps out loud. Our separate guides on analyst, Excel and SQL interview questions cover the question types in detail.
Where the career path leads
After two or three years, analysts usually move in one of four directions: deeper technical work (analytics engineering or data science), deeper business work (a senior analyst owning one function’s metrics), management (leading a reporting or BI team), or a domain move into finance, product or operations roles that value strong analysis. The BLS projects US employment of data scientists to grow 35% from 2025 to 2035, with a median annual wage of $120,230 in May 2025. Those are US figures for a more technical neighbouring occupation, so treat them as a sign of demand for data skills rather than a salary guide for junior analyst roles in your country.
Frequently asked questions
Can I become a data analyst without a degree?
Yes, but expect a longer route. Without a degree you rely more heavily on a portfolio, a structured certificate and an internal move or analyst-adjacent first role such as reporting assistant. Some employers filter on degrees, so apply widely.
How long does it take to become a data analyst?
It depends on your starting point. Someone already doing reporting in another role may move within months by adding SQL and a dashboard tool. A complete beginner should plan for several months of structured study and portfolio work before applying, and often a first role that is analyst-adjacent.
Should I learn Python or SQL first?
SQL first for most people. It appears in more junior analyst adverts, it is quicker to become useful in, and it teaches you how data is structured. Add Python or R once you can query, join and aggregate confidently.
Is a data analyst certificate worth it?
A certificate is worth it as a structured learning path and a CV signal, especially if you are changing careers. It does not replace evidence of applied work, so pair it with portfolio projects you can talk through in an interview.
Sources
- Data Scientists, Occupational Outlook Handbook, U.S. Bureau of Labor Statistics
- Google Data Analytics Certificate, Grow with Google
- Microsoft Certified: Power BI Data Analyst Associate, Microsoft Learn