FAQ: How to get hands-on data experience?
You need experience to get a job in analytics & data science but you can’t get experience without a job - so what can you do?
This question came from a recent 1:1 session.
Analytics and data science aren’t usually entry-level jobs, so if you want to break into the field and don’t have relevant experience, what can you do to get the necessary experience to land a job?
Do “real” projects.
The most impactful experience you can get is paid work, so if you can get your hands on data in your current job, regardless of your job title, start analyzing that data and trying to find insights and make recommendations. Being able to demonstrate that you used data to impact business is the best thing you can talk about in job interviews. (Also, this is how I was able to pivot from marketing to analytics.)
If that’s not an option for you, find other “real” projects. Any of the following are good places to start:
Partner with a local organization that needs help making sense of their data.
If you’re a student or alumni, see if your university has such projects available.
If you have any relationships with a local non-profit or small business, reach out to them.
Reach out to your professors or Ph.D. candidates from your university to see if they need help with their research.
Get involved with organizations that coordinate volunteer projects, such as the National Student Data Corps or DataKind.
Attend “hack nights” where you can work on projects using real data to solve problems. (Something like this.)
Participate in competitions that use real data (like the WiDS annual datathon or check out Driven Data).
Do your own projects.
Doing your own project is more attainable - you don’t have to find an organization that has made its data available or find a competition with upcoming dates, and you also have the freedom to pick whatever subject or dataset you want to work with.
The downside is when you create your own projects, they carry less weight. You don’t have anyone else familiar with the data/subject to review your work and validate that it’s good. You also have to come up with the direction of your project on your own, which can be challenging if you’re inexperienced or lack domain knowledge. You also can’t measure the impact of your work since it was likely never implemented.
A couple of ways to add a little more validity:
Do a project with people from your network, so you can review each other’s work (similar to a peer review, which is common on the job) and learn from each other. If you don’t have anyone like this in your network, join Slack and Discord communities or attend local industry meetup events - some of them have project nights.
Do a Kaggle project that will have a leaderboard and see how you rank. Keep in mind that what makes a Kaggle project achieve a high score isn’t always the same as what makes a “good” project in the business world, but it’s somewhere to start.
Read my answers to past FAQs.
Free Online Events
How to Condense Complex Analyses (Nov 30)
PyLadies Conference (Nov 30 - Dec 3)
PyData Global Virtual Conference (Pay What You Can) (Dec 6-8)
Data Analytics and Generative AI (Dec 13)
Guiding Newcomers to Data Science (Dec 14)
Job Openings
Upgrade to a paid subscription and join my private Discord to find out about openings like these sooner.
Intern, Business Intelligence & Analytics at Workato (Remote)
Junior Data Scientist at Lloyd’s List Intelligence (London, England)
Assistant Data Analyst with the Metropolitan Police (London, England)
More sources for jobs: DataAnalyst.com, BusinessAnalyst.com, Otta.com, my list of companies that hire interns/new grads.