What a data analyst
portfolio should include.
A data analyst portfolio should lead with two or three end-to-end projects, each stating the question you were asked, the data you pulled, the analysis you ran, and the decision it changed, then link the SQL, the notebook, and a screenshot of the dashboard. Recruiters skim for named tools first, so put SQL, Python or R, and your BI tool where they can be seen, and never publish a real employer's raw data or a customer record. Below is the full list of what to put in, the terms an analytics recruiter actually searches, and which of the Portfolio designs suit a work that has to show charts and code side by side.
The sections a data analyst portfolio needs.
An analyst is hired on evidence of reasoning, not on a list of tools, so the portfolio is built around projects that show a question turning into a decision. Work through these, and read the flagged block before you publish a single query.
Three end-to-end projects
Each project should read as a short case: the business question, where the data came from, the cleaning and joins you did, the analysis, and the number that moved because of it. Two strong projects beat six half-finished ones, so cut the weak entries.
SQL you actually wrote
Show a real query with a window function, a CTE, or a non-trivial join, and explain what it answered. SQL is the one skill nearly every analytics posting screens for, so make yours visible rather than implied by a bullet point.
A dashboard, with a link
Embed or screenshot a Tableau, Power BI, or Looker dashboard and say what a stakeholder does with it. A dashboard proves you can turn a query into something a non-technical reader acts on, which is most of the job.
A notebook or scripted analysis
Link a Python or R notebook that walks from raw data to a chart, with comments a hiring manager can follow. Pandas, a regression, or a cohort analysis shows you can go past drag-and-drop when the question needs it.
An experiment or A/B test
If you have run one, describe the hypothesis, the metric, the sample, and how you read significance. Analysts who can design and interpret a test, not just report on one, are the ones who get pulled into product decisions.
Data storytelling
Write a two-paragraph summary of one project for a reader who will never open the query. The ability to say what the data means, and what to do about it, in plain language is what separates a senior analyst from a report generator.
Never include: an employer's real or confidential data
No customer records, no raw exports from a job, no internal revenue figures, no personally identifiable information, and nothing under an NDA. Publishing a former employer's data to a public page is a fireable breach and, with personal data in it, a legal one.
Use a public dataset, a synthetic one you generated, or aggregated numbers with the source named. "Analysed the public NYC taxi dataset to model demand" is safe. Reposting last quarter's real sales table is not, however anonymous it feels.
Terms an analytics recruiter searches.
Analytics recruiters filter their applicant tracking system by named tools and methods. If these are genuinely true of you, spell them out, because the system indexes the exact string, not the skill you meant to signal.
Paste your resume into the free ATS score checker with a real analyst posting to see which of these terms the posting uses and your resume is missing.
Which designs suit a data portfolio.
A data analyst portfolio has to hold charts, code blocks, and prose on the same page without either dominating. Of the 60 Portfolio designs and 48 resume layouts, these are the shapes that fit.
Pick a design that gives each project a card with a chart thumbnail and a short write-up. You need visuals, but not a full-bleed gallery, so choose one built for case studies rather than for photography.
Of the 48 resume layouts, use a single-column one so a two-column skills sidebar does not scramble when a parser reads it. Keep the tools list in its own clearly labelled section.
Order each project so the business question is the headline and the result is the payoff. A reviewer wants to see you think from problem to decision, not admire a chart with no context.
Use a design with a readable monospace block so a SQL snippet stays scannable, and keep chart colours simple. Clarity signals the same taste a hiring manager wants in your dashboards.
Who a data analyst portfolio is not for.
A portfolio helps some analysts and does little for others. Read this before you spend a weekend on it, because for some paths a clean resume and a strong GitHub do more.
Worth building if you
- +Are early in your career or self-taught and need to prove you can do the work without a brand-name employer behind you.
- +Are moving from a nearby field, finance, ops, or academia, into analytics and want to show applied projects.
- +Freelance or consult and want one link that shows range across SQL, BI, and analysis.
- +Want to stand out for a competitive role where every applicant lists the same tools.
Skip it, for now, if you
- −Are a senior analyst applying through referrals, where your track record and a conversation carry more than a public site.
- −Would only be able to fill it with employer data you cannot legally publish. Build public-dataset projects first.
- −Already have a well-organised GitHub or Kaggle profile that recruiters in your niche check by default.
- −Have a deadline this week. Make your resume machine-readable first, then build the portfolio after.
Questions data analysts ask.
Straight answers on projects, data you can use, and whether the effort pays off.
How many projects should a data analyst portfolio have?
Two to four finished projects, each shown end to end from question to decision. Depth beats volume: a reviewer would rather see one project where you cleaned messy data, ran a real analysis, and explained the result than six that stop at a bar chart. Cut anything you cannot speak to in an interview.
Can I use data from my current job?
No, not the raw data and not confidential figures. Use a public dataset, generate a synthetic one, or present only aggregated numbers whose source you name. Publishing an employer's real data exposes you to a breach of contract and, if it contains personal information, to legal liability. When in doubt, rebuild the project on open data.
Do I need to show code, or just dashboards?
Show both. A dashboard proves you can communicate to a stakeholder, and a SQL query or notebook proves you can get to the numbers yourself. Most postings screen for SQL specifically, so make at least one real query visible rather than leaving it implied by a resume bullet.
Should I include a Kaggle competition?
Only if you can frame it as a decision, not a leaderboard rank. Recruiters care less about your score than about how you reasoned: what features you built, how you validated, and what you would do differently. A well-explained mid-table entry can read stronger than a medal with no writeup.
Which BI tool should I feature?
The one the roles you want ask for. Tableau and Power BI dominate most analyst postings, so lead with whichever you know, and mention Looker if your target companies run it. One dashboard you can explain in depth beats three tools listed with nothing behind them.
Where to go next.
Build the site, test your resume, or read how the paste-a-resume flow works.
Turn your analyst
resume into a site.
Paste your resume and Portfolio drafts a clean, project-led website in about a minute. SQL and dashboards up top, no employer data anywhere, published to your own domain with TLS handled for you.