Data analyst course: learn the tools by using them
Excel, SQL, a dashboard tool and Python, in the order analysts usually learn them. You practice in working copies of each tool, every answer is checked, and the path ends in projects and the data analyst interview.
Counts cover every course named in the path, alternatives included.
The path, step by step
Take the steps in this order. Each course starts with a free unit, so you can try a step before you commit to it.
1
Spreadsheets
Most analyst work still starts in a spreadsheet: a file someone emails you, a check before a meeting, a summary for a manager. Learn it in Excel, or in Google Sheets if that is what your team uses.
Clean an export: text that looks like numbers, stray spaces, blanks and duplicates
Look values up with XLOOKUP or INDEX and MATCH, and total by condition with SUMIFS
Summarize with PivotTables and charts that answer one question
Filter and reshape with dynamic arrays such as FILTER, UNIQUE and SORT
2
SQL
SQL is how you get data out of a database yourself, instead of waiting for someone to send an export. The data analyst interview here has a SQL round, and so do many real ones.
Filter, sort and summarize with WHERE, GROUP BY and HAVING
Join tables, and notice when a join multiplies your rows
Rank rows, compare them with the previous row and keep running totals with window functions
The everyday analyst patterns: deduplication, sessions, funnels, cohort retention and month over month
3
A dashboard tool
Pick the course that the jobs you want ask for. The ads usually name it. Both courses go from a first chart to a finished dashboard that someone else can use.
Load CSV and Excel files and take a first look at a table
Filter, group, merge and reshape with groupby, merge and pivot_table
Fix missing values, wrong types and messy text
Resample time series and chart the result
Want only the units this job uses, with your progress across all of them and a timed assessment at the end? Follow the data analyst pathway
Some analyst jobs ask for one more tool. Power Query prepares data that refreshes on its own, BigQuery is a cloud warehouse you query with SQL, Alteryx builds data workflows without code, R is the language many research, health and public sector teams analyze data in, and Google Analytics 4 is where marketing and ecommerce teams read their website data. Add the one your target jobs name.
Each project starts from a fictional company's brief and moves between tools the way analyst work does. The tables you build in one stage are where the next stage starts. Every brief is free to read.
Hearthline
ExExcelSQLSQLpdpandasBIPower BI
ProjectFree
Why is profit falling when sales are up?
Your first week as Hearthline's data analyst. Turn a messy order export into gross profit in Excel, roll it up in SQL, measure what each discount code costs in pandas, chart the margin in Power BI and tell your manager what to change.
You've joined Greenbasket's two-person analytics team. Review the summer promotions, find the stores running out of stock, then explain a bad month to the CFO.
Marketing ran eight promotions this summer and wants to plan the fall. Turn the offers into numbers, measure each one against the weeks before it, total the extra sales and write the memo.
Store managers say shelves keep running empty, but the shelf checks are messy. Mark every check, compare the stores fairly and show availability next to how much was checked.
The monthly report shows September down on August, and the CFO wants to know why before the board meeting. Separate the calendar from a real decline and say where it came from.
Pennant loses too many subscribers in their first three months. Build a churn table without peeking at the future, find who leaves, fit a first model and plan what the retention team should try.
A checkout test offered half the visitors an annual plan. The dashboard says it worked; count the test fairly and decide whether the difference is real before it ships.
Tradewind takes a flat 12% on every sale. The marketplace lead wants tiered fees. Measure what sells in each price band, how far below the asking price it sells, and model what the tiers would earn.
Copperline's on-time report is a spreadsheet someone rebuilds every Monday. Rebuild it on the warehouse: one current row per shipment in Snowflake, a tested dbt model, and a Tableau view operations can trust.
Operations reports one on-time rate and finance reports a lower one, for the same month. Find out why, write the measure both teams agree on, and recommend which number goes to customers.
The courses add 81 projects and case studies of their own, each checked step by step in one tool. See data analytics projects
The interview
The Interview Hub runs the interview round by round. You answer out loud or in writing, the interviewer follows up, and you get a written report that quotes your answers.
Screen
Who you are, why this role, and a short walk through one piece of your work.
SQL
Write and run a query, explain what one row of the result means, and test the awkward rows.
Excel
Talk through a workbook: its formulas, how you checked the source and how it refreshes.
Metrics and experiments
Define a metric, explain why it moved, and decide what an A/B test result supports.
Business case
Structure a business question, show the arithmetic and recommend what to do next.
Fit, review and panel
Questions about how you work, a review of a dashboard, a final panel and practice for the offer.
Project briefs and interview practice. Every project brief is free to read. With a free account you also get 60 minutes of text interviews a month.
The rest of the path. One Course opens one course in full. All Access opens every course on this page, every project stage and spoken interview minutes. See plans and prices
The practice
Each lesson explains one skill in a few short paragraphs, works an example, then hands you four to six exercises in the tool. You finish an exercise and it’s checked straight away. If the answer is off, the message names the likely mistake: a join that doubled your rows, an average that counted blanks as zero, a filter on the wrong date column.
Exercises in a lesson get harder as you go. The first one walks you through the steps. The last one is a small task on new data, the way it would arrive at work. Each unit ends in a practice set that mixes what you have learned, and the checker tries every answer again on a second copy of the data, so a result typed in by hand doesn’t pass.
Certificates and records
The Work Gym is practice. You won’t get a certificate from us, and we won’t claim that a course here equals one from Google or Microsoft. What you will have:
Projects and case studies with a verified record each, collected on one page you can link from a CV.
Answers you have already said out loud, round by round, in the data analyst interview.
Hours in the tools that job ads name, on the kinds of data analysts get handed.
If you want a named credential as well, take a structured certificate for the paper and practice the tools here. We compare the well-known options in the best data analyst courses, compared.
A week on the path
One way to spend about five hours a week:
Two sessions of lessons in the current step.
One session on that unit’s practice set or a case study.
At the end of each step, one project stage that uses the tool, or a text interview round on it.
When a step feels easy, move on. The projects mix tools, so you keep using SQL while you learn pandas.
If a job application includes an aptitude assessment, add the aptitude course to rehearse numerical, verbal and logical reasoning. The regional graduate-hiring guide helps you read the invitation’s rules before choosing a timed mock. This preparation is separate from the analyst’s technical interview.
Marketing analysis
For a role that owns paid acquisition, add Google Ads practice after the measurement foundations. Review campaign goals and search terms before explaining cost and conversion results. The marketing analyst interview provides separate role practice.
Common questions
Is there a certificate at the end?
No. The Work Gym doesn't issue certificates, and this path isn't a program from Google, Microsoft or any other vendor. What you get is a record for each project, case study and mock exam you finish, with a code anyone can check. The Power BI and Tableau courses also have timed mock exams in the format of those vendors' certification exams, if you plan to sit one.
Do I need any experience?
No. Every course starts from first steps: your first formula, your first SELECT, your first chart. If you already know a tool, skip ahead; each course page lists every lesson.
How long does it take?
The lesson estimates add up to about 159 hours if you take one tool at each step. How fast you get through them depends on your week. If you already use Excel every day, the first step goes quickly.
Do I need to install anything?
No. The spreadsheet, the SQL workspace, the dashboard builders and the Python notebook all run in your browser. Power BI Desktop itself runs only on Windows, so the Power BI course is also a way to practice it on a Mac.
Are these the real tools?
They are our own simulators, built to look and behave like the real tools, so the skills transfer to the job. The Work Gym isn't affiliated with Microsoft, Google, Salesforce or Alteryx.
Excel or Google Sheets? Power BI or Tableau?
Read the job ads you would apply to and take the tool they name. If you can't tell, start with Excel and Power BI. Power Query works the same way in both, so what you learn in one carries over to the other.
What does it cost?
The first unit of every course is free, with no account. After that, One Course opens one course and All Access opens all of them, with the project stages and spoken interview time. The pricing page shows every plan and its price.
Will this get me a job?
Nobody can promise that, and we don't. The path gives you practice in the tools analyst jobs ask for, projects you can talk about in detail, and rehearsal for the interview. The applications are up to you.