Selling Yourself in Data Science
From resume to portfolio and beyond
While it might seem data science is all about programming and building fancy machine learning models, when it comes to getting hired as a data scientist, it's not just about what you know but also what you show.
This was a theme in recent video and blog series where I discussed how to sell yourself as a data scientist better. In this article, I synthesize key ideas from this series and share actionable tips to help you level up your data science pitch to potential employers.
Why do I need to sell myself?
Whether you are looking for a full-time role, freelancing, or starting a business, at some point, you will need to sell yourself. While this may offend the idea that only your qualification and experience should matter when seeking work, these factors are moot if you cannot effectively convey them to your audience.
Thus, in many ways, sales is just effective communication.
So instead of focusing on sales hacks and gimmicks, the focus here is on effectively communicating a point (e.g. you’re a great fit for a role) to a particular audience (e.g. the hiring manager). We start with the most obvious place for getting work, your resume.
Resume
Before diving into my dos and don’ts for making a data science resume, I think it’s important to start with the motivational question:
What is the goal of a resume?
Although a resume may have many goals, the one I find the most helpful is:
The goal of a resume is to get an interview.
The reason I think this philosophy is helpful is that people often feel they need to dump all their experience and past work into a resume in an effort to seem more qualified. Ironically, all this tends to do is confuse your audience (the hiring manager) and make you a less compelling candidate.
This brings me to the foundational principle of communication (and thus sales in this context), which is less is more. This principle informs my dos and don’ts for making a data science resume.
My Dos
- Include bullet points with quantifiable impact — In a wall of text, numbers stand out (e.g. 42). Additionally, filtering past work based on impact helps ensure it is both impressive and relevant.
- Add technical skills section at the top — Don’t make a hiring manager hunt for these skills. If they are too busy looking for specific technical skills, they will completely miss everything else you bring to the table.
- Bold key skills in bullet points — This is another way to help those key technical skills stand out.
My Don’ts
- Don’t include an objective section — Your audience only has a finite amount of attention to give your resume, so don’t waste it on some obligatory objectives section. Save that for the interview.
- Don’t make it long (1 page is usually enough) — This helps ensure you only include those details most relevant to the role and your audience.
For a deeper dive into this, there is a YouTube video where I break down the $100,000+ Data Science Resume that got me hired. There you can see a real-world example of these tips in action.
Portfolio
Given resumes are a standard part of any hiring process, it can be difficult to stand out as a candidate from a resume alone. This is where a portfolio is helpful.
A website portfolio helps you stand out as a candidate and gives you credibility as a professional. Additionally, it gives potential employers a dedicated space to explore your background and expertise.
While this all sounds great, if you’re a data scientist like me, then web development is probably not your forte. Lucky for us, however, GitHub has a (no code) solution to this via GitHub Pages, which is a free and built-in function that lets you host a website from a GitHub repository. I break down how to do this in the 2 resources linked below.
🔗 How to Make DS Portfolio with GitHub Pages 📰 | YouTube Video 🎥
Now that you can’t use excessive cost or lack of web dev experience as an excuse let’s turn to the contents of your portfolio. Rather than focusing on what you should do, it often sticks better if we instead focus on what not to do.
This brings up 5 Data Science Portfolio Mistakes that will guarantee your portfolio never gets you hired. These mistakes are summarized below.
- Mistake 1: Don’t Make a Portfolio (this shouldn’t be you since you know about GitHub Pages) — Although making a website has never been easier, most people still do not have a website portfolio. This means just by having one, you immediately distinguish yourself from most candidates.
- Mistake 2: Include Irrelevant Projects — If you are looking for data science roles, then only include projects that showcase those skills. This may sound obvious, but given the variety of technical backgrounds coming into data science, there is a tendency for people to include projects more relevant to software engineering, web development, etc. While these skills can be helpful in DS roles, they are not essential and therefore are a distraction to your audience.
- Mistake 3: Cram in as many Kaggle projects as possible — 1 real-world project > 10 toy projects. While there is serious data science work happening through Kaggle, most often, these projects lack essential elements of real-world data science projects e.g. understanding the business problem, working within existing processes, and synthesizing data from multiple sources.
- Mistake 4: Don’t Include Any Visuals — The key advantage of a portfolio over a resume is the opportunity to include engaging visuals to show your work. Don’t just describe your project. Show it with a short video demo. An example of this is shown in the article linked below.
- Mistake 5: Ignore All Non-technical Aspects — Although technical skills are essential to data science roles, non-technical skills such as communication and domain knowledge are also critical. A simple way to convey these in your portfolio is to make your portfolio pretty and digestible, as well as list out all the domains in which you’ve worked.
For a more detailed discussion on these portfolio mistakes, check out the article on the subject linked below.
I did all this, and still nothing.
While a solid resume and portfolio can get you far, these still may not take you all the way. This is especially true for those who are just getting started in data science.
If this is the case for you, many times, what’s happening is you need more experience and development as a data scientist. This brings up a frustrating loop in which many beginners find themselves, which is something like: “I need experience to get job, but need a job to get experience”.
While things may seem grim, this loop is not necessarily true. Meaning there are ways to get experience outside a job and ways to get a job without experience.
Toward that end, there are at least 5 things anyone can do to become a more attractive data science candidate outside of work experience.
- Education — This is probably the most obvious, yet it is effective. In fact, most data scientists I’ve met have some sort of additional training beyond an undergraduate degree.
- Independent Projects — A cost-free way to learn is to pursue independent projects you find interesting. One thing to avoid here, however, is pursuing a toy project, as mentioned in Data Science portfolio Mistake #3.
- Content Creation — Make content about data science topics and projects. I find this is not only a great way to learn but also a great way to show your skills and competence.
- LinkedIn (networking) — Simply being more active on LinkedIn (posting 3–5x a week) is a great way to increase your exposure to peers and potential employers. The key point here is you may not necessarily need to develop your skills further. Just simply convey the skills you already have more broadly.
- Freelancing — This is a personal favorite for me. Even if you don’t get any contracts, the pace of freelance contracts is much faster than full-time roles. This means you get a lot more reps fine-tuning your resume, writing cover letters, and doing interviews freelancing than otherwise.
A more detailed discussion of these points can be found in the article linked below.
Bonus: A custom email signature
As a final bonus, I’d suggest creating a custom email signature for the account you are using to interact with potential employers. While this may sound like a trivial part of the process, it can go a long way.
As an individual applying for jobs or interacting with clients, you often don’t have a big brand name to back you up. This is where a good-looking professional email signature can make a difference. It is an easy way to give yourself more credibility and share key website links with anyone you email.
If you are struggling to customize your email signature (like me not too long ago), I break down how to do this without spending any money in a recent YouTube video.
