How I Doubled My Income with Data Science and Machine Learning
Many career opportunities exist in the ever-expanding domain of data. Finding your place -- and finding your salary -- is largely up to your dedication, focus, and drive to learn. If you are an aspiring Data Scientist or have already started your professional journey, there are multiple strategies for maximizing your earning potential.
I just want to preface by saying that this article serves as more of a reflection of what I did to get to where I am today. I’m not suggesting that you will achieve the same thing by following the same steps, but I figured that this might provide you with a unique perspective that you might not have thought about before.
With that said, let’s dive into it!
There are three main things that I did to double my income:
- Upskilling (in data science and machine learning)
- Blogging (about data science and machine learning)
- Freelancing (data science and machine learning projects)
Many people today tend to look towards high-risk, high-reward investments like Bitcoin in an attempt to “get rich quick,” but as Warren Buffett said, the best investment you can make is in yourself. This is especially true for those who don’t have money to invest but want to better their finances.
By simply upskilling myself and learning data science and machine learning, I increased my salary by 40% in one year.
There were three main areas that I focused on over the past year:
Data Manipulation (SQL/Pandas)
In my opinion, data manipulation with SQL and Pandas was the most important area that reaped me the most benefits. From my experiences, the most amount of time is spent querying data, exploring data, and wrangling data, all of which require SQL and Pandas. Of all of my data-related jobs (growth marketing analyst, data analyst, data scientist), SQL has been the common denominator and is arguably the most important skill for a data professional.
Here are the resources that I used to self-learn SQL and Pandas:
- Mode’s SQL Tutorial: I always recommend this as a resource because it serves as a great guideline for what concepts to learn, even if you don’t use it to actually learn the concepts. It’s also great because they break down the concepts into different levels of difficulty.
- Mode’s SQL case studies: These case studies are great because they’ll allow you to apply what you’ve learned and critically think about how you would approach real-world scenarios.
- LeetCode database questions: I always use this as a resource, especially when looking for new jobs. It’s a great way to mimic coding interviews that involve SQL.
- Pandas practice problems: Pandas’ syntax is not very intuitive (at least not like SQL), and it never really clicked in my head until I came across this repo of practice problems for Pandas!
I started with Python because of school, and I’ll probably stick with Python for the rest of my life. It’s so far ahead in terms of open source contributions, and it’s straightforward to learn.
There are two main resources that I highly recommend to developing your skills in Python (aside from doing side projects):
- LeetCode algorithms problems: Similarly to SQL, I used LeetCode to learn how to write (somewhat) efficient Python scripts for various problems.
- Tech with Tim: Tech with Tim is a YouTube channel, meaning it’s free, and yet it’s better than most paid courses and boot camps out there. I highly recommend going through his videos and following along with him.
But of course, what fun is being a data scientist without learning machine learning! Below are the two most important resources that I used at the beginning of my career.
- Kaggle’s Intro to Machine Learning: If you were like me and don’t really understand how machine learning exactly works, or you don’t really know how it’s implemented in code, then I highly recommend that you go through this.
- StatQuest: StatQuest is amazing for understanding how machine learning models work. Once you understand the theory, implementing it in code will be simple.
If you want to get an idea of the various machine learning algorithms out there, check out my article here.
I actually wrote a 52-week curriculum that covers SQL, Pandas, Python, and machine learning, which you can check out here.
Now that I covered what I upskilled myself in, you’re probably wondering how I approached this, which is what I’m about to talk about next.
2) Data science and machine learning blogging
Some of you might know that I started a personal initiative called “52 weeks of data science and machine learning,” where I learned, coded, and wrote about something related to data science and machine learning every week for an entire year. This was mainly so that I could keep myself accountable for learning something new on a consistent basis.
After writing over 100 articles and building a follower base of over 20,000 readers, writing now makes up around 25% of my total income.
Here are three tips that gave me a lot of success:
Tip #1: Find the intersection of what you’re good at writing about, what you like to write about, and what people like to read.
Image created by Author.
This is the first tip that I always give to aspiring writers. Ideally, you want to find a niche that satisfies all three of these things.
If you find something that you’re good at writing about and you also like writing about it, but people don’t like to read about it, then you won’t build a follower base (assuming you care about this).
If you find a topic that you’re good at writing about and people like to read about it, but you don’t enjoy writing about it, then you won’t last long as you’ll lose interest.
Lastly, if you find a topic that you enjoy writing about and people like to read about it, but you’re not good at writing about it (because you don’t have enough expertise, for example), then you might not get any traction.
And so, spend the beginning of your journey figuring out your niche. I’m going to elaborate on this in Tip #3.
Tip #2: Understand the mechanics of the platform you’re writing on.
Whether you’re using Medium, Substack, Patreon, or another platform for blogging, make sure that you take the time to understand how the platform works.
I can’t get into too much detail about this, but understanding things like earnings are calculated, how the platform can help you with advertising yourself, and things like that are important things to think about.
By understanding the mechanics of Medium and how it works, I was able to maximize my outreach and ultimately grow my follower base much faster.
The next tip will help you achieve Tips #1 and #2:
Tip #3: Consider the concept “Exploitation vs. Exploration” when creating content.
In order to find the intersection of the three things in Tip #1 and to understand the mechanics of the platform you’re writing on, consider the concept of exploitation vs. exploration.
This idea comes from a statistical problem called “the multi-armed bandit problem.” I won’t get into too much detail, but the main idea behind “exploration and exploitation” is deciding whether to explore and find new potential ideas or to exploit ideas that you already know that works.
At the beginning of your writing/blogging career, it’s in your best interest to explore and try as many ideas as possible to see what works best for you. This means writing about different topics, publishing on different publications, and potentially trying new writing styles.
As you become more developed in your writing style and preferences, you may stumble upon a “recipe” that gives you consistent success in your writing. This is when you can start to exploit that breakthrough and double down on your secret formula.
To summarize, explore as much as possible early in your journey, and as you start to define yourself and find success, start to exploit those insights and ideas that make you successful.
3) Freelancing Projects
The remaining bit of my income came from freelancing projects related to data science and machine learning. The projects that I undertook included written technical papers, writing marketing content, and building models.
When I first started out, I only made around minimum wage from freelancing projects. This made sense because I didn’t have much experience, and I also didn’t know what I was worth. However, by the end of the year, I was able to charge upwards of $50 an hour.
Most of my income came from repeat clients in the tech industry. I actually didn’t have to reach out to anyone else either — I was able to get my clients’ attention through my data science and machine learning blog, and I think that’s the key point for this article.
My data science and machine learning blog didn’t only help me learn on a consistent basis, but it also helped me build my own follower base and helped me get several clients for freelancing.
- A Guide On How To Become A Data Scientist (Step By Step Approach)
- 6 side hustles for an aspiring data scientist
- Data careers are NOT one-size fits all! Tips for uncovering your ideal role in the data space