Portfolio for Data Analysts: What to Include

For a data analyst, the portfolio is not a nice-to-have — it is the interview. A resume can list SQL, Excel and Power BI, but only a portfolio shows whether you can actually take a messy dataset and turn it into a decision. That final step — turning data into a recommendation a business can act on — is the whole job, and it is exactly what a good portfolio demonstrates and a bad one hides. This guide covers what to include, how to present it, and how to make yours read like an analyst's work rather than a student's exercise.

The single biggest differentiator is this: most fresher portfolios stop at a chart. Yours should start with a question and end with an answer. The chart is the middle, not the point.

Who this is for and what recruiters want

You are a fresher or career switcher into data analytics with some SQL, spreadsheet and BI-tool skill and a few projects, or the intent to build them. What a hiring manager wants to see is not a gallery of pretty visuals; it is evidence that you can frame a question, wrangle real data, analyse it soundly, and communicate the result to non-technical people. Every project you include should prove some part of that chain. If it only proves you can make a bar chart, it is not pulling its weight.

The anatomy of a strong analyst project

Each portfolio project should walk through the full workflow, not just the finish line. Present it as a short case study:

DATA ANALYST PROJECT STRUCTURE (per project)
1. THE QUESTION   The business-style question you set out to answer
                    "Which product categories drive the most revenue,
                     and how does that shift by season?"
2. THE DATA       Source, size, and what each field means (public data)
3. CLEANING       What was messy and how you fixed it (nulls, dupes,
                    types) — show a few real steps
4. ANALYSIS       Your SQL queries / Excel work / calculations, visible
5. VISUALISATION   The dashboard or charts that answer the question
6. THE FINDING    Plain-language conclusion + a recommendation
                    "Revenue concentrates in 3 categories; Q3 dip is
                     driven by one region — investigate supply there."

The first and last steps are where freshers stand out, because almost no one does them well. Stating the question up front frames the whole project as purposeful. Ending with a recommendation in plain business language proves you understand that analysis exists to inform decisions, not to admire charts.

Which projects to include

Aim for three to four projects with deliberate variety, so the set as a whole demonstrates range:

A BALANCED FRESHER ANALYST PORTFOLIO
Project 1  SQL-heavy analysis    Querying a relational dataset to answer
                                  layered questions (joins, aggregation)
Project 2  BI dashboard          An interactive Power BI / Tableau
                                  dashboard telling a clear story
Project 3  End-to-end case study Raw data -> cleaning -> analysis ->
                                  visuals -> written recommendation
Project 4  (optional) Excel      Modelling / pivot analysis showing
                                  spreadsheet depth

Variety matters because different roles weight different tools. A retail analytics role may care most about the dashboard; a reporting role may care about SQL. Covering the spread lets one portfolio speak to several job descriptions.

Show the messy middle

The strongest signal in an analyst portfolio is visible, honest analysis — the queries you wrote, the cleaning you did, the assumptions you made. A polished dashboard with no working shown behind it invites suspicion that you followed a tutorial without understanding it. Include your SQL. Note how you handled nulls, duplicates and inconsistent formats. Explain a judgement call you made. The messy middle is where you prove you can actually do the job, so do not hide it behind a pretty front end.

How to present and host it

The container matters less than the content, but presentation still counts. Many analysts index their work on GitHub with a clear README that links to each project's write-up, notebook, and dashboard. Others build a simple site or publish dashboards on a BI platform's public gallery. Whatever you choose, make navigation effortless: a recruiter should reach any project's question, analysis and conclusion in a couple of clicks. Because a GitHub-hosted portfolio leans heavily on its README to orient visitors, invest in that file — the guide to writing a good README for your projects applies directly to analyst project repos and index pages.

Writing up each project

Treat each write-up as a story with a beginning, middle and end. Open with the question and why it matters. Walk through the data and your cleaning. Show the analysis. Present the visual. Close with the finding and what you would recommend or investigate next. Keep the language plain — imagine a business manager, not a professor, reading it. This narrative framing is what turns a collection of charts into a portfolio that reads like the work of someone ready to be hired.

Common mistakes

The recurring ones: portfolios that are all dashboards and no analysis; projects with no stated question, so the work feels aimless; hiding the SQL and cleaning steps; using the same overused sample dataset everyone uses without adding original insight; and stopping at the chart instead of stating a recommendation. Another is neglecting communication — an analyst who cannot explain a finding in a sentence will not pass an interview, and the portfolio is your first chance to prove you can. Finally, never use confidential workplace data; stick to public sources.

Fitting into your wider profile

Your analyst portfolio is one piece of a consistent online presence that also includes your resume and LinkedIn. The projects here should match the skills your profile claims, so a recruiter cross-checking you sees one coherent analyst. Developers building code-heavy portfolios face the same alignment challenge from a different angle — the guide to building a developer portfolio website covers the shared principles of structure, working links and honest project cards that apply to analysts too.

If you want a set of genuine, end-to-end analytics projects — cleaning, SQL, dashboards and written findings — that is exactly what building through a structured program produces. CodeBegun's Data Analytics course in Madhapur, Hyderabad, has students working real datasets into presentable case studies, with placement support to turn that portfolio into interviews. A free counselling session can help you plan projects that show insight, not just charts.