Everyone says "build a portfolio," but I’m struggling with what that actually looks like for a US-based Data Analyst role. Should I focus on complex Python scripts, or are clean Tableau dashboards more important? I have no degree, so this is my only way to show I can handle data cleaning and exploratory data analysis. Does anyone have examples of projects that successfully got them an interview? I’m specifically interested in how to document my SQL queries on GitHub effectively.
3 answers
The best portfolios I see don't just show code; they tell a story. For a US market role, you should have at least three distinct projects. One should focus on Data Cleaning (show your SQL or Python skills in handling "dirty" data), another on Exploratory Data Analysis (EDA), and a final one on Visualization. Use a README file on GitHub to explain the "Business Problem," your "Approach," and the "Result." Recruiters love seeing the "Impact"—for example, "Identified a 10% waste in marketing spend." This shows you think like a business analyst, not just a coder.
Are you using real-world datasets from places like Kaggle or the US Census Bureau, or just the standard "Titanic" dataset? Most hiring managers will ignore projects that use overused datasets because they want to see how you handle unique, messy data
Focus on the "Why" behind your data. I put my best project at the very top of my resume with a direct link. I used SQL to query a database and then visualized it in Power BI.
Keeping the link at the top is a pro tip. Hiring managers only spend seconds on a resume, so making the portfolio accessible is key to proving your skills quickly.
Sean, I've been using local city open data portals. For instance, I analyzed Chicago’s 311 service requests to find trends in neighborhood maintenance. It feels much more authentic and gives me something unique to talk about.