How to Become a Data Analyst From a Non-Technical Background

Much of a data analyst’s work is office work you may already do: reconciling spreadsheets, fixing a messy list, turning a column of numbers into a chart for a manager. In the European Commission’s ESCO classification, data analysts import, inspect, clean, transform, validate, model or interpret collections of data with an organization’s business goals in mind, and may present the results as graphs, charts and dashboards.¹

How to become a data analyst from a non-technical background comes down to adding tools in a sensible order and proving, on real questions, that you can use them:

1. Get fluent in spreadsheets
2. Learn SQL to pull your own data
3. Pick one BI tool for dashboards
4. Learn the statistics analysts use most
5. Add Python or R once the basics feel routine
6. Build two or three projects from your own field
7. Take on data tasks in your current job
8. Apply where your old field counts

If you are still choosing between fields, start with how to plan a career change from the first idea to the first offer. This guide assumes data analysis is the target.

What data analysts do all week

The core of a data analyst’s week is finding, cleaning and summarizing data so that someone else can make a decision, according to official job descriptions from the US Department of Labor, the European Commission and England’s apprenticeship body. The analysis serves a question another person owns.

O*NET, the US occupational database, has no occupation called data analyst: a search for the term returns data scientists and business intelligence analysts as the closest matches.² The business intelligence analyst profile, whose sample job titles include business analyst and BI analyst, covers much of the same work. Its tasks include producing standard or custom reports for executives, managers and clients, maintaining dashboards, and analyzing trends that carry implications for business strategy.³

England’s apprenticeship standard for data analysts gives everyday examples: breaking down retail sales figures to advise on product placement, studying staff retention to shape recruitment, and comparing hospital wait times across departments.⁴ Each starts with a manager’s question and ends with a recommendation, and the spreadsheet or query sits in the middle.

Knowing the field behind the numbers matters for that reason. An analyst in a hospital who knows how wards admit patients is better placed to question a suspicious wait-time figure than a stronger programmer who has never worked there. Before you study anything, read ten job ads in your target sector and underline the verbs, such as report, clean, build and present, then note which of them you already do in some form.

What skills do you need to become a data analyst?

To become a data analyst you need a query language, a spreadsheet, a visualization tool, enough statistics to read data honestly and the ability to explain results to people who do not work with data. ESCO lists statistics, query languages, data cleansing and visual presentation techniques among the essential skills for the occupation.¹

Myth

You have to learn to code before you can do any data analysis.

Fact

ESCO's essential skills for data analysts include data cleansing, query languages, statistics and visual presentation, and name no general programming language.

Beyond the tools, England’s standard expects analysts to gather requirements from clients and colleagues and to present results to a range of stakeholdersstakeholder: In project management, anyone who can affect a project, be affected by it, or believe they are affected by it: an individual, a group or an organization. That includes users, suppliers and colleagues whose work changes, not only the people paying.Full entry in the glossary, alongside knowledge of statistics and of descriptive, predictive and prescriptive analytics.⁴ The US Bureau of Labor Statistics names communication among the important qualities for data scientists, a related profile in its Occupational Outlook Handbook: they must explain results to technical and nontechnical audiences.⁵

A fundraiser at a charity shows how the pieces fit. Pulling last year’s donations by region is the query, removing duplicate donors is the cleaning, the bar chart is the visual, and explaining to trustees why one region fell is the communication. Each skill on the official lists maps to one part of a task like that, so learn them in the order a task uses them.

Job ads show which tools employers name most often. In US postings for business intelligence analysts, SQL led all software skills named in 2025 job ads.⁶

The study

Moderate evidence

What US job ads for BI analysts asked for in 2025

O*NET counted how often each software skill appeared in 2025 US job postings for business intelligence analysts, using data from the labor-market company Lightcast. SQL appeared in about 35% of postings, Microsoft Power BI and Python in about 20% each, Tableau in 19%, Microsoft Excel in 17% and R in 10%.⁶

Read these as signals of demand. A posting records what an employer hoped to find, and who got hired may differ; the data also covers only one US occupation code. A low share for Excel may simply mean employers take spreadsheets for granted; the data cannot say. Elsewhere, ESCO’s skill lists for the EU and job boards in your own country are the places to check what employers near you ask for.

Breaking in without analyst experience or a degree

Getting a first analyst job without analyst experience is possible but hard: official US sources describe this family of jobs as usually needing a degree and considerable experience. O*NET places business intelligence analysts in Job Zone Four, where most occupations require a bachelor’s degree, though some do not, and workers usually need several years of work-related experience or training.³

Official US projections point to growth in a related occupation. The BLS projects that employment of US data scientists will grow 35 percent from 2025 to 2035, much faster than average, driven by demand for data-driven decisions and the spread of AI systems.⁵ That figure covers a more technical occupation in one country, so it says little about how many people compete for each junior analyst opening. No official source we checked counts applicants per entry-level analyst job.

Do you need a degree? For data scientists, the BLS says the typical entry route is a bachelor’s degree, in fields such as mathematics, statistics, computer science, business or engineering.⁵ Many employers have dropped degree requirements from ads, yet in a report published in February 2024, Harvard Business School researchers and the Burning Glass Institute estimated from US career histories that the change reached not even 1 in 700 US hires the year before.⁷ The estimate comes from company research without peer review. Without a degree, plan on proof of skill doing more of the work, and expect some ads to screen you out.

A career changer’s edge is the field they come from. A logistics coordinator applying for analyst roles at shipping firms competes on knowledge most new graduates lack.

How to become a data analyst, step by step

The order below mirrors beginner course outlines such as Google’s Data Analytics certificate, which moves from preparing and cleaning data through analysis and visualization to an introduction to Python.⁸ No study has compared learning orders for career changers, so treat this one as a sensible default.

1. Spreadsheets: sorting, filtering, lookups and cleaning; ESCO lists data cleansing as an essential skill
2. SQL: pulling and joining your own data; ESCO lists query languages as essential
3. One BI tool: Power BI or Tableau, whichever your target ads name, for dashboards
4. Statistics: averages, medians, spread and the limits of a correlation
5. Python or R: for automating repeated work once the basics feel routine

1. Get fluent in spreadsheets

Spreadsheets are the easiest place to learn the habits of analysis, because most office workers already use one: sorting, filtering, lookups, pivot tables and charts. Cleaning comes first, and a payroll officer who has hunted down a duplicated employee ID has already done it.

2. Learn SQL to pull your own data

SQL is the language for asking questions of a database. Learning it means you stop waiting for someone to export a file for you. Start with selecting, filtering, grouping and joining two tables. England’s standard expects analysts to work across different data structures and database designs, which in practice includes joining tables.⁴

3. Pick one BI tool for dashboards

A business intelligence (BI) tool turns queries into dashboards that managers can open themselves. Choose Power BI or Tableau by checking which one your target employers’ ads name. Microsoft’s Power BI certification describes the analyst’s work as preparing, modeling, visualizing and analyzing data, which is a useful syllabus even if you never sit the exam.⁹

4. Learn the statistics analysts use most

Much of the statistics analysts use day to day is descriptive: averages, the medianmedian: The middle value when numbers are put in order, so half fall below it and half above. Unlike the average (mean), it is not pulled up or down by a few very large or very small values, which is why it is often used for incomes and wealth.Full entry in the glossary, spread, percentages and comparisons over time. Restraint matters as much as method. A correlationcorrelation: A measure of how closely two things move together, running from minus one, where one rises as the other falls, through zero, meaning no link, to plus one. It says how strong the relationship is, not what causes it, and it is not a percentage.Full entry in the glossary between training hours and sales does not show that training raised sales, and a manager will act on what your chart seems to say.

5. Add Python or R once the basics feel routine

Python or R lets you automate repeated cleaning and handle data too large for a spreadsheet. The BLS says data scientists must learn data-oriented programming languages.⁵ Learn one, and learn it on a task you already do by hand, such as merging twelve monthly files.

6. Build two or three projects from your own field

A portfolio project answers one real question from start to finish: the data, the cleaning, the query, a chart and a short recommendation.

1. 1A questionone a manager in your field would ask
2. 2Datapublic or made up, never personal
3. 3Clean and queryspreadsheet or SQL
4. 4Chartone clear picture
5. 5Recommendationwhat to do, in two sentences

Draw the questions from the field you know. A teacher might study attendance against term dates in public education data; a retail supervisor might compare sales by weekday and weather. We found no study of which portfolios persuade employers, so this is practitioner advice.

Never put your employer’s data, or anyone’s personal data, in a public project. England’s standard asks analysts to handle data in line with laws such as GDPRGDPR: The General Data Protection Regulation, the European Union's main data protection law, applied since 2018. It sets rules for how organizations collect and use personal data about people in the EU, including organizations based elsewhere, and gives people rights over that data.Full entry in the glossary, and employers elsewhere have their own rules; public or made-up datasets avoid the problem.⁴

7. Take on data tasks in your current job

Your current job is the cheapest place to gain experience that counts. Offer to build the monthly report, clean the customer list or set up a team dashboard. In England, apprentices can be new or current employees, so an employer can train existing staff through the Level 4 data analyst apprenticeship.¹⁰ In England, each apprenticeship standard sets its own minimum of paid off-the-job training, according to GOV.UK guidance updated in October 2026; for the data analyst standard, Skills England lists 370 hours.¹¹⁴ Elsewhere, ask HR whether your employer funds courses; how tuition reimbursement works covers what to check with HR before a course starts.

8. Apply where your old field counts

Search under several titles. O*NET’s sample titles for this kind of work include business intelligence analyst and market intelligence analyst.³ England’s standard lists junior analyst, marketing data analyst and departmental data analyst among typical titles.⁴ Describe your old job in the new field’s terms, using a method for spotting transferable skills in past work. Lead your CV with projects and the data tasks from your current job.

From first spreadsheet to first application

• Read ten job ads in your target sector and underline the tools and verbs
• Clean and summarize one messy spreadsheet from public data
• Write SQL queries that group and join two tables
• Build one dashboard in the BI tool your target ads name
• Finish two or three projects that answer questions from your field
• Ask for one recurring data task at your current job
• Apply under several analyst titles in the sector you know

What training asks of you in months and money

No official source measures how long it takes a career changer to land a first analyst job, and course timelines are only a floor. Google’s certificate page estimates 6 months at 10 hours a week to finish its program, while England’s Level 4 data analyst apprenticeship typically lasts 24 months, alongside paid work.⁸⁴ Those figures assume a steady pace; at half the hours a week, expect the same course to take about twice as long, and job hunting comes on top.

Prices differ between countries and are revised often, so check the current figure on the provider’s own page. As of October 2026, Coursera charges 49 US dollars a month in the US and Canada for Google’s certificate after a 7-day free trial, and says the cost may be lower in other countries where the certificate is offered.⁸ Microsoft prices its Power BI exam by the country where you sit it.⁹ Before you pay for any course, check its refund terms and whether your employer would fund it.

The job ads in your target sector are the best guide to which paid credential, if any, employers there ask for.

The bottom line

For many career changers, the most practical route into data analysis runs through the field they already know: learn spreadsheets and SQL first, then build two or three projects that answer real questions from your current work. This week, pick one recurring report in your job and offer to rebuild it with cleaner data and a clear chart.