What the job data says to learn first
O*NET's profile for Business Intelligence Analysts (15-2051.01), the closest standard occupation to many data analyst job titles, lists producing standard or custom reports for stakeholders and maintaining dashboards and databases among its top tasks. In its technology skills, Microsoft Excel and Microsoft Power BI are flagged as in demand, and several SQL tools appear (Transact-SQL, Oracle PL/SQL, SQL Server Reporting Services). O*NET places the role in Job Zone Four, and 68% of surveyed respondents said a bachelor's degree is required.
The Bureau of Labor Statistics profiles the more advanced Data Scientists occupation: median pay was $120,230 in May 2025, and employment is projected to grow 35% from 2025 to 2035. BLS describes the work as identifying data sources, collecting and analyzing data, building models and presenting findings with visualization software, and says the typical entry route is a bachelor's degree. Treat that as the ceiling of the path, not the entry point: the analyst stages below are the foundation those roles build on.
The order matters because each stage feeds the next. Spreadsheets teach you what tidy rows and columns look like; SQL lets you pull those rows from a database; statistics tells you whether a difference is real; visualization turns it into a report someone reads; Python handles the jobs that outgrow a spreadsheet.
The six stages, with a free and a paid option each
Hours in the third column are our planning ranges for someone starting from zero; where a provider publishes hours, the option column shows them. Pick one option per stage. Doing both the free and the paid option for the same stage is usually wasted time.
| Stage | What you should be able to do | Planning estimate | Free option | Paid option |
|---|---|---|---|---|
| 1. Spreadsheets | Sort, filter, IF, VLOOKUP, PivotTables, basic charts | 15 to 25 hours | Coursera Guided Project: Introduction to Data Analysis using Microsoft Excel (about 2 hours, intermediate level, listed as free in the catalog) | Google Data Analytics courses 3 and 4 (19 and 16 hours) on preparing and cleaning data |
| 2. SQL | SELECT, WHERE, ORDER BY, JOINs, GROUP BY and aggregates, NULL handling | 20 to 30 hours | SQLBolt: interactive in-browser lessons from SELECT through JOINs, aggregates and NULLs (no price is stated on its homepage) | SQL for Data Science, UC Davis on Coursera (4 modules, about 15 hours of module time, included with Coursera Plus) |
| 3. Statistics | Summaries, distributions, sampling, confidence intervals, hypothesis tests, linear regression | 30 to 45 hours | OpenIntro Statistics, free PDF (9 chapters from intro to data through multiple and logistic regression) | OpenIntro print copy ($25 black and white, $40 colour); paid courses are optional here |
| 4. Visualization and BI | Choose a chart for the question, build a dashboard, explain it | 15 to 25 hours | Power BI Desktop, which Microsoft describes as free for building reports (Microsoft points to Power BI Pro, listed at $14 per user per month paid yearly, for publishing and sharing reports) | Google Data Analytics course 6, Share Data Through the Art of Visualization (19 hours, uses Tableau) |
| 5. One language (Python) | Variables, loops, functions, then pandas for tables and a plotting library | 30 to 90 hours | Educative Learn Python 3, free (about 10 hours, 17 quizzes, in-browser editor) | Educative Become a Data Analyst path (80 hours, 224 lessons: pandas, seaborn, projects) or Google course 7, Introduction to Data Analysis Using Python (27 hours) |
| 6. Portfolio | Two or three original analyses with a question, cleaned data, findings and limits | 20 to 40 hours | Your own project on a public dataset, published as a write-up | Google Data Analytics capstone case study (11 hours) |
One paid program or a free stack?
The Google Data Analytics Professional Certificate covers most of the stages in one sequence: nine courses whose listed hours add up to 152, which Coursera frames as about 6 months at 10 hours a week. It is beginner level with no prerequisites and is included with Coursera Plus. The certificate page's FAQ quotes $49 a month in the U.S. and Canada after a 7-day trial; the Coursera Plus page lists $59 a month or $399 a year. Check which subscription checkout is actually offering before you pay.
The gap in that program is statistics: its course list does not include a dedicated statistics course, so pair it with OpenIntro chapters 1 to 8. The gap in a free stack is cohesion and graded feedback: you assemble the sequence yourself and nobody marks your work.
- Choose one paid program if you need deadlines and a single checklist, and can work 8 to 10 hours a week so the subscription months stay few.
- Choose the free stack if you are cost-sensitive, already self-directed, or want to spend more time on statistics and SQL than a certificate requires.
- Skip paid Python courses at first if spreadsheets and SQL are not yet comfortable; on O*NET's BI analyst profile the in-demand flags go to Excel and Power BI, with SQL tools such as Transact-SQL and Oracle PL/SQL marked as hot technologies.
Worked example: a 26-week plan at 8 hours a week (illustrative)
This is a planning example, not a provider schedule. It totals about 208 hours, inside the ranges above.
| Weeks | Stage | Hours | Done when |
|---|---|---|---|
| 1 to 3 | Spreadsheets | 24 | You can build a PivotTable summary from a raw export and explain each column |
| 4 to 7 | SQL | 32 | You can write a three-table JOIN with GROUP BY without notes |
| 8 to 12 | Statistics | 40 | You can say whether two group averages differ and how sure you are |
| 13 to 15 | Visualization | 24 | One one-page dashboard answering a stated question |
| 16 to 22 | Python | 56 | You can load a CSV in pandas, clean it, group it and plot it |
| 23 to 26 | Portfolio | 32 | Two published write-ups using your own question and dataset |
Stage checklist before you move on
Self-testing before moving on is deliberate: practice testing is one of the two techniques rated high utility in Dunlosky and colleagues' 2013 review. See our retrieval practice and spaced practice guides for how to schedule these checks.
- Spreadsheets: rebuild one PivotTable from scratch a week after first learning it, without the video.
- SQL: solve five new questions on a dataset you have not seen before; JOINs and GROUP BY are the usual stall points.
- Statistics: explain a confidence interval in two sentences to someone outside the field.
- Visualization: for each chart, write the question it answers above it; remove any chart without one.
- Python: reproduce one spreadsheet analysis in pandas and check the totals match.
- Portfolio: every project states its question, data source, cleaning steps, findings and limitations.
Questions before you decide
In what order should I learn data analyst skills?
Spreadsheets, then SQL, then basic statistics, then a visualization tool, then one programming language such as Python, and finally an original portfolio project. O*NET lists Excel and Power BI as in-demand tools and several SQL tools for business intelligence analysts.
How long does the Google Data Analytics certificate take?
Coursera describes it as about 6 months at 10 hours a week. Its nine listed courses add up to 152 hours, including a 27-hour Python course and an 11-hour capstone case study. It is beginner level with no prerequisites.
Can I learn data analysis for free?
Yes, stage by stage: OpenIntro Statistics offers a free PDF, Educative's Learn Python 3 course is free, Microsoft describes Power BI Desktop as free for building reports, and some Coursera Guided Projects on Excel are listed as free. Paying adds structure, graded work and a certificate.
Sources & verification
Product details and prices can change. Check the linked provider before buying.
- O*NET OnLine: Business Intelligence Analysts (15-2051.01) Accessed 30 Oct 2026
- U.S. Bureau of Labor Statistics: Occupational Outlook Handbook, Data Scientists Accessed 30 Oct 2026
- Coursera: Google Data Analytics Professional Certificate Accessed 30 Oct 2026
- Coursera: SQL for Data Science (UC Davis) Accessed 30 Oct 2026
- Coursera Plus pricing page Accessed 30 Oct 2026
- Educative: Learn Python 3, free interactive course page Accessed 30 Oct 2026
- Educative: Become a Data Analyst skill path Accessed 30 Oct 2026
- OpenIntro: OpenIntro Statistics textbook page Accessed 30 Oct 2026
- SQLBolt: interactive SQL lessons Accessed 30 Oct 2026
- Microsoft: Power BI Desktop product page Accessed 30 Oct 2026
- Coursera Guided Project: Introduction to Data Analysis using Microsoft Excel Accessed 30 Oct 2026
- Coursera: Guided Projects catalog listing Accessed 30 Oct 2026
- Dunlosky, Rawson, Marsh, Nathan and Willingham (2013), Psychological Science in the Public Interest 14(1):4-58, DOI 10.1177/1529100612453266 Accessed 30 Oct 2026
Sources link directly to providers. Product buttons may use separately labeled affiliate links. Read our disclosure.