Where to Find Datasets for Your Data Analytics Projects (And What to Do With Them)
- Sarah Rajani
- 3 days ago
- 7 min read
Updated: 5 hours ago
A Practical Guide to Finding (and Using) Datasets for Data Analytics Projects
By now, you've probably heard about the importance of having a data analytics portfolio.
In my previous two posts, I shared how to write up your data projects and how to build a portfolio site to house them.
But how do you even choose a dataset to begin with?
I think that might be the hardest part of the process.
How do you pick a dataset worth building a project around? Is there a particular type of dataset you need?
The answer is not so straightforward.
Actually, a lot of it comes down to your preferences: what you like, what tools you plan on working with, and the industry you’re hoping to work in.
I have done data projects that run the gamut of topics, but I only include a small selection in my portfolio, which you can take a look at on my site. But narrowing down the choices was easier said than done.
That's why I wanted to write this post. I break down the various options for choosing your dataset, from open-source data you can explore on your own, to guided projects that walk you through a realistic use case.
How to Choose a Dataset for Your Project
Before you even start looking at datasets, you should think about the type of project you want to work on. It's much easier to look for something when you can narrow down the options.
Here are a few things to think about:
What are You Interested in?
It’s easier (and more fun) to work on something you care about. If you're into music, sports, or fashion, there’s data out there for that. Don’t force yourself into a topic just because it seems “impressive.” You’ll get more out of a project you enjoy.
If you don't like sports, don't waste your time building a project looking at basketball stats. You'll lose interest and motivation quickly.
What Tools Do You Want to Practice?
Pick a dataset that works well with the tool you’re learning
For example:
If you’re practicing SQL, look for structured data in rows and columns.
If you're learning Power BI or Tableau, get something you can slice into charts and visuals.
If you want to get better at Excel, find a dataset with fewer rows and cleaner formatting.
This step is not necessary, especially if you're not sure exactly what to to look for. It just makes it easier if you choose data that works well with your tool.
What Industry Do You Want to Work in?
If you're targeting a role in finance, healthcare, government, or retail, then choose datasets that mirror the problems in that industry. That way, your project becomes a more relevant portfolio piece.
One of my favorite projects was on crime stats in my city, because I was interested in roles within the criminal and justice sectors. Now I work in that space, and I think the project gave me that edge I needed.
How Much Time Do You Have?
Some datasets are large and messy. Others are clean and manageable. If you're working on something small for a weekend portfolio build, start with a guided project or a clean CSV. Save the multi-table mess for when you’re ready to work on your data wrangling skills.
Once you’ve got answers to these questions, choosing a dataset gets way easier.
Free and Open-Source Datasets
Open-source datasets are the best type when you want full control. What I mean by that is you can choose the dataset, clean it, analyze it, and present the insights in any way you want. No guidelines, no structure, just data for you to work with in whichever way you choose. The sky is the limit here.
They are great for:
Portfolio projects
Practicing real-world data wrangling
Showing you can ask and answer business questions independently
Kaggle
Kaggle is one of the most popular platforms for finding datasets. There are so many options available, from sports to healthcare to e-commerce.
Pros:
Huge variety, strong community, and often comes with example notebooks.
Some datasets are raw, others are curated.
You can also join competitions.
Example project:
Use the Netflix dataset to analyze genre trends or build a recommendation dashboard.
Data.gov
This one has over 300,000 datasets from the U.S. government across topics like education, agriculture, transportation, and public safety.
Example project:
You can analyze Los Angeles crime statistics and visualize trends over time.
Canada Open Data
If you’re in Canada (like me!), this one has some interesting options. Topics range from health and transportation to energy and demographics.
Example project:
Compare urban vs rural commute times, or look at how transit usage has changed post-COVID.
European Data Portal
Great for comparing economic, environmental, and demographic data across EU countries.
Example project:
Analyze renewable energy adoption trends over time.
NYC Taxi & Limousine Commission Data
This one allows you to access raw trip data from yellow taxis, green cabs, Uber, and Lyft.
Example project:
Identify the busiest pickup zones and times, and build a Power BI heat map.
NASA Earth Data
if you like earth science, you can find environmental and satellite data for geospatial or climate-related analysis. There's data for atmosphere, agriculture, the oceans, and so much more.
Example project:
Track temperature anomalies or visualize CO2 levels over time.
FiveThirtyEight Datasets
Clean, well-documented datasets from the team behind FiveThirtyEight’s articles. The datasets run the gamut from politics, sports, culture, and so much more.
Example project:
Analyze tennis match times or NBA game stats.
UN Data
Global stats on health, education, population, and development.
Example project:
Explore literacy rates and educational access by region.
UCI Machine Learning Repository
If you're interested in data science, this is a good one for structured datasets used in machine learning, but great for regular data analytics too.
Example project:
Use the wine quality dataset to find which features are most associated with higher ratings.
10. Google Dataset Search
Think of this like Google for datasets. You can find niche or industry-specific data fast.
Example project:
Search for open datasets on food delivery trends, resale markets, or electric vehicles.
11. Datahub.io
Smaller but clean collection of datasets focused on finance, health, and economics.
Example project:
Visualize changes in world GDP.
Guided Projects (Don’t Skip These)
Guided projects are great ways to learn, and they are exactly what you need when you are starting out. Otherwise, how else can you learn? You need to have someone walk you through the process. Then you can work on your own projects and answer your own questions.
A guided project can help you:
Learn structure.
Practice new tools.
Understand how to frame and solve problems.
So here's how you do it:
Follow the project step by step.
Then re-do it on your own without the guide.
Add new questions in your re-do project, build new visuals, write your own summary.
Here are some places you can get guided projects:
Google Data Analytics Capstone (Coursera)
You get a fictional bike-share dataset and have to analyze customer behavior. They provide guidelines but the actual project is up to you, so there is a lot of learning involved with this.
Example project:
Compare casual vs annual riders. Recommend strategies to increase paid subscriptions.
DataCamp Projects
DataCamp's platforms has tons of mini projects with clear prompts, walkthroughs, and code templates.
Example project:
Explore NYC 311 service complaints. Slice by borough, type, or time.
Maven Analytics Challenges
You get the dataset and a business question in their monthly challenge, which is great for practice and feedback.
Example project:
Pizza sales dashboard: When are sales highest? What should staffing look like?
Dataquest Projects
Browser-based guided projects where you analyze, clean, and visualize within their interface.
Example project:
Explore Hacker News data to see what titles drive the most comments.
YouTube
And let's not forget YouTube. I've found some fantastic guided projects on there. You just have to type in whatever too you want and then "guided project."
For example: "SQL Server Guided project"
What Questions Should You Ask?
You don’t need to analyze everything. You just need to ask a few good questions. Think of some interesting things someone at that company might want to know.
Here are types of questions you can start with:
Type | What it Means | Example |
|---|---|---|
Descriptive | What happened? | What were the top 5 selling products last month? |
Diagnostic | Why did it happen? | Why did churn increase in Q1? |
Comparative | How do two things differ? | Are new users more active than returning users? |
Exploratory | What patterns or trends exist? | Are there seasonal spikes in customer support calls? |
Predictive | What will likely happen next? | Can we forecast next month’s website traffic? |
Operational | What actions should we take? | What days need more staff based on foot traffic? |
Stick with descriptive or comparative questions at first, as they’re easier to explain and more practical for general business questions.
So, Where Do You Start?
Start with one project. Any project. It doesn’t have to be groundbreaking, it just has to be something you put yourself into. Incorporate your skillset, investigative and analytical strengths, and writing abilities.
Pick a dataset.
Choose one tool you’re learning.
Ask a question you’re curious about.
That’s it.
Don’t worry about making it perfect. No one has a perfect project writeup, especially not their first one.
As you gain more experience, you can come back to your older projects, tweak them, and build on them. Or, you can pick something new and do it again.
That’s how you build your data skillset and a portfolio that reflects how you think.
Once you've selected the dataset you want to use, or the guided project you want to walk through, you're ready to turn it into something you can showcase.
These two guides will walk you through what to do next:
👉 How to Write About Your Data Projects – Learn how to structure your findings, explain your process, and tell a clear data story.
👉 How to Build a Data Analyst Portfolio – See what to include, where to host it, and how to make yours stand out.
Tips for Selecting the Right Dataset
Align with Your Interests:
Choose datasets that resonate with your personal or professional interests to keep you motivated.
Assess Data Quality:
Ensure the dataset is clean, well-documented, and relevant to your project objectives.
Start Small:
If you're a beginner, opt for smaller datasets to manage complexity and focus on honing your skills.
Check Licensing: Always review the dataset's licensing terms to ensure you're compliant with usage rights.
Final Tips
Finding the right dataset is a crucial step in your data analytics journey. By using the resources listed above, you can access a ton of information to practice your skills, build your portfolio, and gain insights into data you might encounter in a real business scenario.
Remember, the best projects are the ones that interest you and that you're excited to explore and analyze.
Enjoyed this post? Share your thoughts in the comments.
I write about data, career transitions, and making analytics easier to understand.
Subscribe to my newsletter to get the latest updates.
And come say hi on LinkedIn. I’d love to see what you’re working on.
Exactly what I needed to know. Thank you.