How to become a data analyst in 2026: the 6-step path and the best course for each step

The skill. Before SQL and dashboards, an analyst knows how to turn a vague business question into a precise one, clean a messy file and summarize it with a pivot table. It is also the skill you can use on Monday with your team's own spreadsheets.

The main course. Google's certificate on Coursera has 9 courses. It covers spreadsheets, SQL, Python and Tableau, plus new modules on using AI to clean and visualize data. Google estimates about 6 months at 10 hours a week, self-paced, with no degree or experience required. It costs $49 a month in the US and Canada after a 7-day free trial, and Google says most learners finish for less than $300. The certificate carries an ACE recommendation for up to 12 college credits, and Google runs a consortium of more than 150 US employers. With 4.8 out of 5 from over 182,000 reviews and almost 4 million enrollments, nothing else on this page comes close in reach.

Who it is for. The beginner who wants structure, a recognizable name on the résumé and a low price before committing thousands to a bootcamp.

Its limit. Assessments are mostly multiple choice, so the certificate proves you passed its quizzes, not that you can do the work. Practitioners repeat that employers care more about solid projects with real data than about the certificate. Google reports that 75% of graduates saw a positive career outcome within six months, but that figure comes from Google's own survey. One detail to watch: the page still lists R among the tools while its FAQ says R is no longer covered.

The alternative. If spreadsheets are your real gap, LinkedIn Learning's “Learning Excel: Data Analysis” course runs just over 3 hours and is free if your employer already licenses the platform. Maven Analytics (step 6) also teaches Excel with a business focus.

The tip. Treat the capstone as your first portfolio piece, not as homework. The certificate is only worth it if you take the projects seriously and use real data, ideally from your own job.

The skill. SQL, the language used to query databases, is the filter. In threads where hiring managers discuss what made a candidate stand out, “demonstrated SQL skills” comes up again and again. Without SQL, you wait for someone else to pull your numbers.

The main course. DataCamp pairs a learning track with a real exam. The Associate Data Analyst in SQL track takes about 39 hours of interactive exercises. The Associate certification then asks you to pass a timed SQL exam of 2 hours and a practical exam on a real dataset. The full Data Analyst certification adds a 1-hour exam in SQL plus R or Python, and a recorded presentation of your findings. You get two attempts per exam and 30 days to finish, and the credential stays valid for two years. The certification is included in Premium, which costs $28 a month billed annually.

Who it is for. The analyst who needs proof of SQL that someone actually graded, and the manager who wants to track a team on one platform. The Teams plan costs $28 per user per month, billed annually, for 2 users and up.

Its limit. The credential is less known than Google's or Microsoft's, so pair it with projects. Both prices assume a yearly commitment.

The alternative. On a zero budget, try the free 19-hour data analyst bootcamp by Alex The Analyst, published on freeCodeCamp. It covers SQL, Excel pivots and XLOOKUP, Tableau, Power BI with DAX and Power Query, and Python with Pandas. There is no exam, but it is enough to decide whether you like the work.

The tip. Write your practice queries against your company's own data, such as a CRM export or a sales table. A query that answers a question your boss actually asked is the best interview story you can bring.

The skill. A dashboard your director opens every Monday that refreshes itself. That takes a BI tool (business intelligence, the software that turns data into dashboards) and its modeling layer: in Power BI, Power Query to prepare data and DAX to calculate.

The main course. Microsoft's data analyst career path on Microsoft Learn is free and self-paced, with an instructor-led option. It leads to the PL-300 certification: a 100-minute proctored exam on preparing, modeling, visualizing and analyzing data, and on managing and securing Power BI. The exam costs $165 in the US, and you need 700 points to pass. Microsoft provides a free practice assessment and an exam sandbox that mimics the real interface. If you fail, you can retake it 24 hours later.

Who it is for. Anyone whose company runs on Microsoft 365, and any manager who wants one shared credential for the whole team. It is Microsoft's own credential for Power BI, the BI tool most courses on this page teach, and any employer can verify it on Microsoft's site. It is also the only credential here that renews for free: a short online assessment every 12 months.

Its limit. It covers Power BI and nothing else: no SQL depth, no Python, no statistics.

The alternative. If your team prefers structured video lessons, Maven Analytics teaches Power BI on the same subscription as its Excel courses (step 6). For other ways to prepare, see our ranking of Power BI certification courses.

The tip. Register for the exam with a personal Microsoft account, not your work account. With a work account, you lose your certification history the day you leave the company. If your company is not on Power BI, compare the tools first in our ranking of BI and data visualization software.

The skill. Python takes over when the spreadsheet freezes and SQL is not enough: automating a weekly report, joining files with millions of rows, calling an API. It is also where analysts run their statistics: knowing whether a sales bump is real or noise, and reading a correlation without calling it a cause.

The main course. IBM's certificate on Coursera has 11 courses and about 180 hours. It puts Excel and Python side by side. The courses run from Excel and SQL to Python with Pandas, NumPy and Jupyter, charts with Plotly, then IBM Cognos Analytics and Tableau. Inside the Python courses, you also cover descriptive statistics and statistical analysis with SciPy. An optional course covers generative AI. IBM estimates 4 months at 10 hours a week. It sits inside Coursera Plus, Coursera offers financial aid, and it carries an ACE recommendation for up to 12 college credits. It holds 4.6 out of 5 from nearly 100,000 reviews.

Who it is for. The analyst who finished the spreadsheet and SQL steps and wants Python plus a first layer of statistics without paying bootcamp prices. It only asks for high-school math.

Its limit. There is no dedicated statistics course, and no inference beyond the basics. It is self-paced only, with no live teaching. Cognos Analytics takes real time, yet the tools that come up in the job requirements we read are Excel, SQL, Power BI and Tableau. The page also contradicts itself on duration: 4 months in one place, as little as 5 months in its FAQ.

The alternative. For a company that wants statistics taught live, General Assembly's Data Analytics Bootcamp has a dedicated block of statistics and math for analysts, next to SQL, Excel, Tableau, Power BI and Python. It runs 12 weeks full-time or 32 weeks part-time (about 15 hours a week), with a team training offer. Tuition is not published: ask for it in writing before you compare. On a budget, DataCamp's Data Analyst in Python track (36 hours) feeds into the certification from step 2.

The tip. Learn Python after SQL, never before. Practitioner roadmaps agree on the order: Excel, SQL, one dashboard tool, then Python. Then test one real change from your job: did last month's campaign actually move the number?

The skill. Turning an analysis into a decision. A correct chart that nobody acts on is wasted work. Storytelling means choosing the one chart that matters, removing the clutter, and saying what you want the audience to do.

The main course. Storytelling with Data runs a half-day live virtual workshop for $595: 4 hours plus an optional office hour, built around a before-and-after method. For more depth, its 8-week online course costs $1,895 (an early-bird rate exists) and takes 3 to 5 hours a week. It includes two books, 6 months of community access and a certificate of completion.

Who it is for. The analyst whose numbers are right but whose presentations do not land, and the team lead who wants everyone to build charts the same way. Custom team training is available, and a group discount kicks in from 5 attendees.

Its limit. It teaches communication, not tools: you will not learn Power BI or SQL here. The certificate of completion carries little weight on its own. Team training is priced on request.

The alternative. If your employer already licenses LinkedIn Learning, the “Become a Data Analyst” path includes Learning Data Visualization among its 12 courses (41 hours in total). It is video only, with no graded project.

The tip. Rebuild one recurring report from your job using the method, then show the before and after to your manager. That single slide often does more to get the next course funded than any certificate.

The skill. Showing what you can do. A hiring manager, or your own boss for an internal move, wants to see two or three analyses on real data: the question, the data, the result and the decision it led to. Outside your company, the interview almost always includes a SQL test.

The main course. Maven Analytics is built around that proof. Its free plan already includes a learning plan, skill assessments and a public portfolio. Pro, at $34 a month or $399 a year, adds badges and credentials. The courses cover Excel, Power BI, SQL and Python with a business focus. Teams costs $39 per seat per month or $468 per seat per year, with a 2-seat minimum. It adds an admin dashboard, assignments and progress reporting; Enterprise starts at 10 seats with custom pricing.

Who it is for. The analyst who wants a public portfolio to send with an application, and the manager who wants to assign courses to a team and see progress. Its few alumni reviews are glowing; one describes going from not knowing what a VLOOKUP is to a certified Excel expert.

Its limit. The review sample is small, and Python and statistics run less deep than at IBM or DataCamp. Badges are Maven's own, not an industry credential.

The alternative. The free Alex The Analyst bootcamp (step 2) ends with portfolio and job-search tips. And the capstone of the Google certificate (step 1) gives you a first project to publish.

The tip. Inside your company, an internal move is often faster than an outside application: take on analyst-style tasks in your current role, document each one as a portfolio piece, then ask for the title.