are you sitting on right now?
10? 50?
Maybe you’ve crossed 100 and you’re starting to wonder if you’ll ever break in.
Well, I’ve been there myself.
I sent over 400 applications before securing my first data science job.
Image by author from my LinkedIn.
However, over the past few years, I have landed $100k+ job offers from companies like Gousto, Deliveroo, DoorDash, Wise, and a couple of startups.
So, in this article, I’m breaking down the exact mistakes I made so you can skip the struggle and fast-track your path to a high-paying data science career.
Let’s get into it!
Useless Learning
The very first thing you need to do to get a data science job is obviously learn some data science.
The problem is that it’s so easy to learn completely useless information that is not actually needed when trying to land a job.
I spent weeks learning topics that I was never asked about or used in any interview process I have been in. And, I have gone through well over 100 interviews at this point.
Things like AWS, Docker, unit testing, etc., rarely come up in interviews. I mean, how can someone really test your AWS knowledge in 1 hour?
Yet many others and I spend time learning these topics even though it’s a complete waste of time if you are looking to get hired as quickly as possible.
I encourage all my coaching clients to stick to studying the fundamentals:
- Probability theory
- The basic supervised and unsupervised learning algorithms
- Easy to medium Leetcode questions
- The steps behind building a machine learning model
- Statistical testing and experimentation
- The fundamental machine learning concepts like gradient descent, bias vs variance and cross validation
These are all areas that always come up in interviews and are where you should invest your time.
You need to ruthlessly prioritise learning the fundamentals, as they pay the most significant dividends in the long run.
Scatter Gun Approach
It shouldn’t take 400 applications to land your first job.
It took me that long because I was using the “scatter gun” strategy. I was spamming LinkedIn’s Easy Apply like there was no tomorrow.
The success rate and numbers speak for themselves: this method resulted in very few interviews.
What I should have done is employ the “sniper” method and hone in on roles where I had a clear advantage.
I reckon you are thinking,
But Egor, I’ve got no advantages
This is simply a myth.
Everyone has an advantage; you just haven’t found it yet.
For example, you can target roles where you have…
- A university thesis relevant to a specific industry.
- Side projects that solve a company’s exact pain points.
- Living in a location with less local competition.
Also, don’t shoot for the moon right off the bat.
You are very unlikely to get a FAANG offer if you have no prior experience, unless you attended an excellent school and were top of your class.
The majority of people should start at smaller companies and slowly work their way up. It’s precisely what I did, and it’s a much more sustainable strategy.
Stop wasting energy on roles you aren’t a fit for and start aiming where you can actually win.
Un-optimised Resume
My first resume completely sucked, like, honestly, it was complete dogwater. I am even surprised that I landed a role in the end.
The truth is, most people think their resume is good. However, I have reviewed hundreds of data science resumes, and most of them are pretty bad.
I have a whole article explaining what a great data science resume looks like, but let me break down the key points here.
- Use a clean template with simple formatting. You can find my one here.
- Keep it to one page, unless you have a decade of relevant experience.
- Always mention metrics, numbers and especially financial impact.
- Lead with your expertise, as that’s what recruiters mainly hire for.
- Use action words like “led”, “developed,” “executed,” and “conducted.” You want to be clear that you did these things.
- Don’t spam too many programming languages and technologies; this is a red flag, since I doubt you know all of them.
It’s honestly not too hard to create a great resume; you need to put in the time, and I am talking at least 10 hours.
It may sound like a lot, but it’s the most crucial document in your professional life, so why try to cheap out on it?
Tailor Your Resume
Every application I submitted used the exact same resume.
The same generic stuff, not personalised at all for the company or role I was going for.
In this market, being generic and basic doesn’t cut it.
What I should have done and what I tell every coaching client I work with is to tailor your resume to every single role you apply for.
Yes, I literally mean every single job.
Look at the job description, identify the key words and phrases, and insert them into your resume.
I know I just said not to spam loads of programming languages and technologies, but that is not what I am suggesting here.
I am asking you to be tactful with the tools you add, so you can explicitly showcase that you have the exact skills the company is after.
You want to optimise your resume as much as possible against the ATS (application tracking system) to avoid any unnecessary auto-rejection.
I know this sounds boring and a lot of work, but this is what you need to do if you want to get a job in today’s competitive market.
Networking & Referrals
If I could give you just one “hack” to land more interviews, it would be to get a referral.
According to this post:
Higher Hiring Success Rates Employee referrals are four times more likely to be hired than applicants who apply through job boards. According to a study by Jobvite, 40% of hires come from referrals, despite referrals making up only 7% of applicants.
The leverage you get with a referral is simply crazy.
We live in an era where people are socially awkward and are so afraid of rejection that those who do have the courage to ask for a referral hold a golden ticket.
You should start with the low-hanging fruit. I am 100% certain that someone in your family or friend group works at a company where they could refer you.
Often, the only thing standing between you and an interview is the simple act of asking.
Please stop reading this immediately and write down 10 of your closest friends or family members, along with where they work.
Then check each company to see if they are hiring for a data scientist role, and ask for a referral.
Sounds simple right? That’s because it is.
If for some reason you are that weird outlier that has no connections whatsoever, which I highly doubt, then you need to actively build your network.
LinkedIn is still criminally underutilised by most job hunters. What other platform gives you access to people who work at companies you want to be at and allows you to interact with them?
When you think about it, it’s crazy powerful.
You should aim for ~50 LinkedIn connection invites per week to people at your target companies.
Make sure you send a thoughtful and personal connection message, but you don’t need to be spending more than 15 minutes per message.
An even better approach is to connect with people you have an “affinity” with, such as those who share your university, hometown, or common interests.
People are much more likely to connect with you if you share similar traits or backgrounds; it’s baseline human psychology.
Build rapport first by asking about their experience; once you’ve established a connection, share your credentials and ask for a referral.
It’s a numbers game, so don’t be discouraged if most people don’t reply.
Always Follow Up
Most people think that after you have applied for a job, your work is finished.
Time to kick up your feet and have a lovely coffee while you wait for a response on your application.
Oh boy, do I wish life were so easy.
If you are doing what everyone else does, you are going to get the same results: very few interviews.
So what else should you do?
After you have submitted your application, find the hiring manager, talent partner, or recruiter linked to that job posting.
You can find their LinkedIn profile or email; it doesn’t really matter.
Then message them something like this:
Hi [name],
I just saw this data scientist role from you guys are I am very interested in applying (or have applied).
I have been working as a Data Scientist and Machine Learning Engineer for over 4 years across insurance, e-commerce and logistics in the the classical ML, pricing models, forecasting and optimisation domains.
I would love to have a conversation about the role!
Let me know if there is anything else I should do.
(Obviously tailor it to yourself!)
What you have just done is put yourself front and centre of their mind for the job. This is somewhere you clearly want to be.
When I have done this in the past, if they reply, you are almost certainly getting an initial interview.
Combine this step with a referral and you are golden for getting an initial interview.
Mock Interviews
Walking into an interview without preparation is like taking a driving test without ever getting behind the wheel.
You’re simply setting yourself up for failure.
Mock interviews are the ultimate “cheat code.” They allow you to over-prepare, which is precisely where you want to be in case some curveballs get thrown your way during the actual interview.
It took me a long time to realise the power of mock interviews. I flopped several early interviews that, in hindsight, I should have easily passed.
Today, I cruise through the process because of the sheer volume of practice I’ve put in.
Because data science and machine learning roles aren’t as standardised as software engineering, the interview process can feel like the “wild west,” with many variations.
To cover your bases, you should run mocks for:
- ML/DS Theory — Testing your foundational knowledge.
- Pair Programming — Live coding under pressure.
- Behavioural — Polishing your “soft skills” and storytelling abilities.
- Case Study Presentations — Communicating technical things in a digestible manner.
This may sound like a lot of work, and it is.
The majority of people think getting a data science job is a walk in the park; that’s why they get zero results and start to “blame the market”, and take no personal accountability.
If you follow the steps in this article, you will eventually land a data science role.
However, if you want to speed run the process, then I invite you to join the Data Science Launchpad.
This is my coaching programme, where you will get direct support from a community of like-minded individuals and me, along with a proven step-by-step framework for landing data science jobs.
You can apply to the Data Science Launchpad using the link below:
https://coaching.egorhowell.com
Another Thing!
Join my free newsletter where I share weekly tips, insights, and advice from my experience as a practising data scientist and machine learning engineer. Plus, as a subscriber, you’ll get my FREE Resume Template!
Dishing The Data
Weekly emails helping you land your first job in data science or machine learningnewsletter.egorhowell.com
