Data Analysis Projects for Your Portfolio: 10 Ideas from Beginner to Advanced

By Skillancy Editorial Team · 20 October 2026 · 5 min read

A cohort retention dashboard with bars and a trend line

The right data analysis projects do more for an analyst job search than another certificate. A good project proves you can take a messy business question, clean real data, analyze it with SQL or Python, and explain a recommendation. This guide gives ten portfolio-ready project ideas, from beginner to advanced, with the business question, the tools and the steps for each, plus a template for presenting projects so hiring managers can review them in minutes.

What makes a project worth showing

A business question, not a dataset. "Which customer segments are churning and why?" is a project. "Analysis of a churn dataset" is a homework exercise.

Real cleaning. Missing values, duplicates, inconsistent labels and date formats. Hiring managers know real data is messy.

A decision at the end. A recommendation, with the evidence and its limits.

Something they can open. A GitHub repository, a public dashboard or a short write-up.

Where to find data

Kaggle datasets: a large public collection across retail, finance, sports and more.

Open Government Data Platform India (data.gov.in): public datasets from Indian government departments.

Company and app exports you are allowed to use: your own spending, fitness or study data, anonymized.

Synthetic data you generate: useful for practicing SQL patterns, as long as you say it is synthetic.

Always check a dataset's license and remove personal information before publishing.

Beginner projects

1. Sales performance dashboard (Excel or Tableau)

Question: Which products, regions and months drive revenue, and where is growth slowing? Clean the data, build a pivot-based summary, then a one-page dashboard with revenue trend, top products and regional comparison. Finish with three bullet-point insights.

2. Cricket performance analysis (SQL)

Question: Which batters perform best in the final overs, and does it change by venue? Load ball-by-ball data, then use GROUP BY, joins and window functions to compare strike rates by phase. If you want a guided start, the free SQL Super Over game teaches the same SQL patterns on IPL-style data.

3. Personal finance tracker (Excel)

Question: Where does monthly spending go, and which categories are rising? Categorize transactions, build a monthly trend with conditional formatting, and set a simple budget-vs-actual view.

Intermediate projects

4. E-commerce funnel analysis (SQL + Tableau)

Question: Where do users drop off between visit, add-to-cart and purchase? Build the funnel by step and by channel, compare conversion rates, and identify the step with the biggest leak. Recommend one experiment to test.

5. Customer cohort retention (SQL or Python)

Question: What share of customers who first bought in a given month return in later months? Assign each customer a cohort month, build a retention table, and visualize it as a heatmap. This project exercises CTEs and window functions; our guide to CTE in SQL covers the patterns.

6. Delivery delay analysis (SQL + Python)

Question: Which routes, warehouses or time slots cause the most late deliveries? Calculate on-time rate by segment, look for patterns by day of week, and quantify the cost of delays. Operations and supply-chain teams hire for exactly this kind of work.

7. Marketing campaign performance (Excel or Python)

Question: Which campaigns delivered the lowest cost per acquisition, and is that difference meaningful? Calculate spend, conversions and cost per acquisition by campaign and channel, and be explicit about what the data can and cannot prove.

Advanced projects

8. A/B test analysis (Python)

Question: Did a new checkout design increase conversion? Check the sample sizes, compute the conversion difference with a confidence interval, and write a recommendation that includes practical significance, not only a p-value.

9. RFM customer segmentation (SQL + Python)

Question: Which customers are loyal, at risk or lost? Score recency, frequency and monetary value, build segments, and suggest a different action for each segment.

10. AI-assisted reporting workflow

Question: Can a weekly report be produced automatically and explained in plain English? Use SQL for the numbers, then an AI assistant or an automation tool to draft the summary, and add a validation step that checks the figures. Our article on data analytics with AI explains where AI helps and where you still need to check it.

How to present each project

Use the same structure for every project so a reviewer can scan quickly:

Section	What to write
Question	The business question in one sentence
Data	Source, size, and what you cleaned
Approach	Tools and the main steps
Findings	Three to five plain-language results
Recommendation	What a business should do, and the caveats
Links	Code, dashboard, write-up

Put this as the README of each GitHub repository, and summarize each project in two lines on your data analyst resume.

Mistakes to avoid

Copying a popular tutorial project unchanged. Reviewers recognize the common ones; add your own question and analysis.

Ten charts and no conclusion.

No explanation of how the data was cleaned.

Claims the data cannot support, such as causation from a correlation.

Broken links.

Frequently asked questions

How many data analysis projects do I need for a portfolio?

Two or three strong, end-to-end projects are usually enough for an entry-level role. Quality and clear explanation matter more than quantity.

Which tools should my projects use?

SQL in at least one project, Excel or a BI tool such as Tableau or Power BI for presentation, and Python if the roles you want ask for it.

Can I use public datasets like Kaggle?

Yes. Make the project your own with a specific business question, real cleaning steps and a clear recommendation.

Where should I host my projects?

GitHub for code and README write-ups, Tableau Public or a similar service for dashboards, and a short LinkedIn post or blog write-up to explain the findings.

Sources and publishing references

Open Government Data Platform India: https://data.gov.in/

Kaggle Datasets: https://www.kaggle.com/datasets