Three Facts You Need to Realize About Doing Data Science Projects
Realize this and let your competition be the one who receives a rejection e-mail
Having a clearly thought out, well-crafted, and professionally executed project will do WONDERS for helping you stand out from all of your competition.
A lot of leaders in the field- people who have actually worked as data scientists — will agree with me that they can even take the place of internship, co-op, or actual work experience.
It will make the difference between being invited for an interview and getting an auto-rejection e-mail.
And this is true whether you’re a career transitioner or been in the game for one or two years.
A bad project, on the other hand, will get you nowhere in this field.
Over the last few years I’ve helped nearly 3,000 data scientists land their first job in data science.
And during that time, I’ve reviewed A LOT of portfolio projects and take-home assignments.
The sad truth is that most of the work I’ve seen is not well done.
I’ve seen great resumes and candidates who look excellent on paper, but after looking at their project I ended up tossing their application in the trash.
Because the way they completed their projects or take-home assignments clearly indicated that they had no clue what they were doing.
They made all the classical mistakes of an amateur: spaghetti code, disorganized repository, no clear methodology, no way for me to understand what it was they were trying to learn or accomplish with the project.
Instead of demonstrating their skill, it demonstrated that this person is someone I simply cannot trust to do good work on my team.
If you want your work to stand out — you need to demonstrate practical knowledge.
You want the quality of your work to separate you from the hundreds or thousands of people vying for that same job.
You need to demonstrate that you’re able to execute with a compass instead of a map.
Here are three concrete reasons doing projects is so important to not only demonstrate your skill as a data scientist, but also to develop the intuition of working like a data scientist so that you can tackle any data problem and not be fearful.
Whether that’s in a portfolio project, on a take-home assignment, or when you’re on the job.
- Learning is not enough
- You need experience to get a job
- You have no principles for working
Let’s dig into these in more detail
Learning is not enough
To quote one of my favorite philosophers, Epictetus:
“That’s why the philosophers warn us not to be satisfied with mere learning, but to add practice and then training. For as time passes we forget what we learned and end up doing the opposite, and hold opinions the opposite of what we should.”
A hiring manager isn’t going to be able to open your head, peek inside your brain, and verify if you know how to do this thing or that, or how well you understand this, that or the other.
In fact, they don’t care about what you have in your head. They care about what you can do for them with what you have in your head.
That’s why you need to do a project. You have to leave an artifact of your ability to produce something tangible with the knowledge in your head.
Every boot camp or university course you’ve taken has given you a disconnected set of tools that you can put into your toolbox — but no course teaches you how to put all of that together to solve problems in the real world.
You need experience using those tools to do something tangible that other people can see.
You need experience to get a job
But that doesn’t mean you need a job to get experience.
We live in a day and age where data is available everywhere.
You don’t need to be in a job at a company to get access to data. If you know where to look and how to search for it, you’ll quickly see that it is everywhere.
Just waiting for you to do something with it.
You may have a naive notion that at work all of your assignments will come to you with the data neatly wrapped up, packaged in a box, with a bow on top.
And you open up a little envelope and it has all of the nice step-by-step instructions telling you with exact specifications what you’re supposed to do.
No.
That’s not how it works.
And the more equipped you are to be able to deal with that ambiguity, the more invaluable and indispensable you are.
You have to be comfortable with ambiguity and problem finding.
You can gain exposure dealing with ambiguity and problem finding by creating a project, taking a vague idea, clarifying the problem statement, and defining the path forward with a set of guiding principles.
You have no principles for working
The data science life cycle is a set of principles for doing data science work.
All problems are different, but you can still use a set of principles to help you find a solution. You can develop and cultivate this intuition through projects.
By doing a well-thought out and executed project you can confidently state in an interview that you understand and have used the data science process.
That even though you don’t have work experience, you still have experience working with and using the data science life cycle. Because you’ve done it in a personal project and have proof to share with them.
And you’re confident that if you’re hired, you’ll be able to do this at work
Doing well constructed projects will teach you how to work as a data scientist.
If you’re brand new to the field, and haven’t had your first job yet then you likely don’t know what it’s like to deliver results and manage your time wisely.
But you can learn how to do that.
You can practice working through the data science lifecycle and you can work in sprints.
A few really well thought out, professionally executed projects, using the data science life cycle after you have completed your schooling and are in the job search can substitute for on the job experience.
I’d love to hear what you think. Leave a comment and let’s talk about it.
And remember my friends: You’ve got one life on this planet, why not try to do something big?
