Reviewed: July 2026
Most technical fields have a cruel problem: your best work is locked behind an employer's firewall, so interviews reduce to describing things nobody can check. Data analytics is the exception. Published dashboards and write-ups can be opened by any hiring manager, in thirty seconds, before they have even replied to your application. That is an unfair advantage over almost every other discipline — and most candidates still waste it by publishing the wrong things.
Here is how to build the portfolio that gets opened, and survives being opened.
What is a portfolio actually for?
Not to prove you can make charts. Software can make charts. A portfolio exists to show three things a resume cannot:
- You can find the question. Real analytics starts with an ambiguous situation, not a clean brief
- You can handle real data. Which means missing values, inconsistent categories, and judgment calls you can defend
- You can change a decision. The output of analysis is someone doing something differently, and your work should visibly aim at that
The bar to aim for: a stranger opens your project, understands the question within a minute, follows your reasoning to the insight, and thinks "I could hand this person a problem." That is the entire game.
Which three projects should you build?
Depth beats breadth, because a hiring manager will open at most two things. Build these three, properly.
1. The public dashboard. One polished, interactive piece on a topic with a genuine question behind it. This is your first impression — the link that goes in your resume header. Design it for a decision, not for density: a clear headline finding, drill-down for the curious, annotations that show thinking.
2. The end-to-end analysis. Start from messy public data and document the whole journey: the question, the cleaning decisions, the SQL, the exploration including dead ends, the finding, and a recommendation with stated limitations. This is the project that demonstrates you are an analyst rather than a dashboard operator — and it is where you show your SQL, which most analyst roles live in daily.
3. The domain bridge. Analyse something from the industry you came from — logistics, retail, healthcare, finance, whatever you know. Career switchers systematically undervalue this one: it converts your past into an advantage no fresh graduate can copy, and it gives you an interview story where you understand the business meaning of every number.
| Project | What it proves | The mistake to avoid |
|---|---|---|
| Public dashboard | Visual craft, design for a decision | Decoration over clarity |
| End-to-end analysis | Process, SQL, honest reasoning | Hiding the messy parts |
| Domain bridge | Business judgment, your unfair advantage | Picking a generic topic instead |
Where should your work live?
Tableau Public is the strongest publishing venue in the field — interactive work, open to anyone with the link, no gatekeeping. If you are Tableau-side, publishing there should be routine; it is precisely why our Tableau course makes a public portfolio piece a graded deliverable rather than a suggestion.
Power BI candidates need one extra step, since day-to-day Power BI work lives inside organisations: publish the thinking instead. A written case study with screenshots, the model design explained, and a short walkthrough video covers everything an interviewer needs. Our Microsoft Power BI course builds exactly these artefacts through its projects.
Wherever the visuals live, add a simple index — a one-page site or even a well-organised GitHub profile — so one link reaches everything. Put that link at the top of your resume, not buried at the bottom, and check it from a phone: a surprising share of first looks happen there, and a dashboard that collapses on a small screen is a first impression too.
Want a practitioner's eyes on your portfolio plan before you build it? That is a normal demo-session conversation. Book a demo session
What separates a portfolio piece from a tutorial artefact?
Hiring managers have seen every tutorial's final dashboard hundreds of times, and they recognise them instantly. The tells of real work:
- You chose the question. Not "here is sales data visualised" but "I wanted to know why X, and here is what the data says"
- The data resisted. Real datasets are messy; showing how you handled that credibly is evidence, not embarrassment
- There is a recommendation. Analysis that ends at "interesting" is unfinished; end at "therefore"
- Limitations are stated. "This data cannot tell us Y" reads as seniority, because it is
- It is annotated. A dashboard that needs you standing next to it explaining is half-done
None of this requires advanced techniques. A straightforward analysis, honestly reasoned and clearly presented, outperforms a complicated one that cannot explain itself.
How should each project be written up?
The write-up is half the artefact, because it is where your reasoning becomes visible — and reasoning is what is being hired. A structure that works for every project:
- The question, in one sentence. Before any chart: what did you want to know, and why would anyone care?
- The data and its problems. Where it came from, what was missing or inconsistent, and the two or three cleaning decisions that required actual judgment — with your justification
- The analysis, including one dead end. The path that did not work signals honesty and method; polished infallibility signals a tutorial
- The finding, stated plainly. One headline insight a non-analyst could repeat in a meeting
- The recommendation and its limits. What should change because of this, and what this data cannot settle
Keep it shorter than you think — the reader is deciding whether to interview you, not grading a thesis. And write the alt text, label the axes, and date the work: small professionalism compounds, and hiring managers in this field notice chart hygiene the way editors notice spelling.
How does the portfolio actually get used in hiring?
Three moments. Screening: a recruiter clicks your link and spends under a minute — your headline project must land its point in that window. The interview: "walk me through this" is now a standard analytics interview segment, and discussing work you actually did is dramatically easier than whiteboard puzzles. The offer conversation: demonstrated judgment is what separates you from equally-certified competition.
Which is why the deeper preparation — the interview itself, the resume, the search strategy — still matters around the portfolio; we covered that side in How to Get Your First IT Job. And if you are still choosing which tool to build on in the first place, our comparison of Power BI and Tableau settles it by the only measure that counts: your target employers.
Start this weekend
Pick the question you are genuinely curious about — there is one — and start project two, the end-to-end analysis. It is the highest-value piece and the one that improves fastest with feedback.
If you would rather build all three with structure, deadlines and a practitioner reviewing your work at professional standard, that is what our analytics programmes are for: Data Analytics for the full analyst toolkit, Tableau or Microsoft Power BI for tool depth with portfolio deliverables built in. Book a demo session and bring the dataset you have in mind.