Data Analytics Portfolio

You've probably heard about data analytics portfolios. Why are they so important? If you're just starting and don't have years of experience, they're your way of showing a future employer the projects you've worked on. Similar to the portfolios that photographers and designers have — it's a way to showcase your skills.

Today, let's talk about what actually belongs in your portfolio.

Common Mistakes

I've reviewed many portfolios. The most common mistakes are:

❌ Projects with no business context

❌ Beautiful notebooks that don't answer "so what?"

❌ Ten half-finished projects instead of three strong ones

I'm sure you've also heard people saying "Don't use generic datasets like Titanic or Iris".

The problem with those datasets isn't that they've been used by so many aspiring data analysts, but that they're very distant from actual work. Of course, you can stretch that — by doing classification of flower types, you can also classify customers. By predicting who would survive on the Titanic, you can predict who would churn as a bank user. But our brains are always looking for the simplest, easiest solution. And if an HR person has 100 other portfolios that are clearly aligned with the business, they will give preference to that option instead of making assumptions about what you can and can't do. So there isn't such a big problem with using public datasets, as long as the project itself is aligned with the job you're applying for.

What Hiring Managers Want to See

When I was hiring analysts at my consultancy, here's what made someone's portfolio stand out:

✅ Clear business questions - Not just "I analysed this data" but "I wanted to understand X to help make decision Y". As we discussed in the last newsletter, you always start an analysis with a specific question in mind

✅ Clean, documented work - Comments explaining your thinking, not just code

✅ End-to-end process - Data cleaning, analysis, visualization, conclusions/recommendations

✅ Honest about limitations - "Here's what I couldn't answer and why"

✅ Executive summary - What was the purpose of the project, what were the findings, next steps

The 3-Project Portfolio Formula

You don't need ten projects. You need three strong ones:

Project 1: End-to-End Business Analysis

Shows you can: Find patterns, clean messy data, ask good questions, visualise insights. Every data analytics project has these components

Project 2: SQL-Heavy Analysis

92% of jobs mention SQL. Prove you can use it.

What to show:

• Complex queries (JOINs, CTEs, window functions)

• Working with multiple tables

• Clear documentation explaining your logic

Project 3: Interactive Dashboard

Show you can present data to non-technical people. It can be an Excel dashboard, Tableau/PowerBI/Looker. It’s more about how you present data rather than the tool you use. I’ve seen overwhelming PowerBI reports, and very sophisticated elegant Excel dashboards.

Executive Summary

The most important part of this whole exercise is the final presentation. You did all the hard work by analysing data, joining tables, making visuals, but that was just 40% of the work. Now you need to put the project together. Imagine you are working on a PowerPoint presentation explaining what you’ve done - that’s what we are aiming for. Not the code, not SQL queries. We are looking for insights supported by data and visuals.

The final step should and will take the longest in both portfolio projects and real work. ‘Selling’ your work takes time.

The Bottom Line

Your portfolio isn't about showing you can write code. It's about showing you can solve business problems with data.

Three strong projects that demonstrate:

• You understand what businesses need

• You can extract and analyse data

• You can communicate insights clearly

That's more valuable than ten half-finished Jupyter notebooks.

Start small. Pick one project. Make it excellent. Then move to the next.

And remember: the perfect portfolio is the one you actually finish and use to apply for jobs.

Done is better than perfect.

Keep pushing 💪

Karina

Introducing: The Complete Business Analysis Portfolio Project

I've spent the last week building a portfolio-ready Data Analysis project that shows you exactly how real analysts work.

The Project:

You'll work on an e-commerce dataset (ShopTrend 2024) — analysing sales trends, product performance and campaign effectiveness.

You'll practice data cleaning, exploratory analysis, visualisation, and storytelling — the skills actually used in analyst roles.

What's Inside:

Python Project Notebook - Step-by-step Jupyter notebook walking through complete e-commerce analysis — from business question to insights and visualisations (Matplotlib, Seaborn, Plotly)

Dataset - Synthetic (generated) dataset with seasonality, missing values, and realistic errors

Dataset Generator - Python script to generate your own custom dataset (adjust rows, time period, categories, data types). Don't want to run Python? I included a ready-to-use AI prompt — just paste into ChatGPT or Claude and download your data

Alternative Datasets - List of other datasets you can use if you want to practice

Notion Portfolio Template - Complete portfolio presentation of the ShopTrend analysis, fully filled out. This shows you exactly how to present your work: how I frame business questions, document findings, structure recommendations, and display visuals. Use it as your reference for building portfolio projects, or customise the template for your own analyses.

Who this is for:

✅ You know Python/Pandas but struggle to start your own projects

✅ You've done courses but don't know how to build something portfolio-worthy

✅ You need a complete project you can show employers

Who this isn't for:

❌ Complete beginners (learn Python basics first)

❌ Data scientists (this is analyst-level work, not ML)

Python Tip

I found some great library - pivottablejs.

It gives you an interactive, drag-and-drop pivot UI inside your notebook.

! It won't show a chart automatically. You need to drag fields into: Rows Columns Values like in Excel, and optionally choose a chart type (e.g., bar, line, heatmap).

#pip install pivottablejs

import pandas as pd
from pivottablejs import pivot_ui

df = pd.DataFrame({
    "Region": ["East", "West", "East", "South", "North"],
    "Product": ["Apple", "Apple", "Banana", "Banana", "Carrot"],
    "Sales": [100, 150, 200, 130, 90]
})

pivot_ui(df)

Grab your freebies if you haven’t done already:

Data Storytelling Guide

Data Playbook (CV template, Books on Data Analytics and Data Science, Examples of portfolio projects)

Need more help?

Just starting with Python? Wondering if programming is for you?

Master key data analysis tasks like cleaning, filtering, pivot and grouping data using Pandas, and learn how to present your insights visually with Matplotlib with ‘Data Analysis with Python’ masterclass.

Grab your Pandas CheatSheet here. Everything you need to know about Pandas - from file operations to visualisations in one place.

Get CV, Portfolio and LinkedIn Review.

More from me: YouTube | TikTok | Instagram | Threads | LinkedIn