🗝️ How to never look for a job again

👋 Hi!

Most weeks, this newsletter brings one practical dataviz tip you can apply right away. You can read past issues and subscribe here.
​
Today I want to tackle something different, because I keep getting the same question: how do you get a job in the data world, and how do you switch from another field?

And even if you already have a job, this still matters. As I wrote recently, you are always looking for your next one.

I am not a recruiter and I would not claim to be a job market expert. But I have worked in academia, in a big tech company, and as a freelancer, and luckily I've never had to apply for a job once in my entire career.

So here is what helped me, and the five steps I would work on if I wanted to enter the data space today again.

​
1️⃣ Specify what you want

Start by understanding the data "landscape".
​
So many jobs are related with the field of data. The 3 main groups are Data engineer, Data analyst and Data scientist. They all work with data but rely on very different tools and skills. I even made a little interactive Viz to illustrate this:
​
​

This big picture is a great start, but there are many more specific roles inside each group.

For example, data analysis covers a lot of ground. Some people crunch DNA datasets (bioinformaticians). Some build advanced statistical models (statisticians). Some focus on crafting clear visuals (dataviz engineers). Some collect and report new datasets daily (data journalists). And that list keeps going.

todo: Before you read the next steps, take a few seconds to make sure you truly understand the exact kind of job you are targeting.

​

​

2️⃣ Learn, with focus

✏️ Figure out what you need.

This field is technical. Soft skills help, but they will not get you through the door on their own.

After step 1 you know the role you want, so now you need to map the skills you need. They vary a lot from one job to another. A data engineer does not exist without strong SQL. A data scientist cannot function without a solid math background. And the pattern continues across every role.

Listing all the required skills and tools is not easy, but the roadmap.sh website helps. It gives you clear diagrams showing what you need to become a data analyst, a data scientist, or a data engineer.
​

todo: take a pen and a paper. List all the skills and tools you need to know to get your dream job
​
​
​⭐️ Become the expert at one thing

Here comes the tricky part: there is no chance you will learn everything on that list. Nobody does.
​
Instead, I suggest you become solid across the board, but become truly exceptional in one clear area. That is what makes you stand out.

If you want to be a data analyst, get comfortable with R, statistics, and dataviz. Then push one speciality to a very high level, whether it is Shiny dashboards, Bayesian methods, time series forecasting or whatever you love.

When I worked at Datadog, one person was the clear CSS expert. Another one was the go to person for Typescript. The respect they had was incredible. They could land any job they wanted because they were known for being world class in one narrow field.

todo: Before you move on, open a job board and search for the speciality you chose. Those are the roles you want to target!
​
​
​🤓 Learn, learn learn

Watch videos, read books, take a bootcamp. We all learn differently, so there is no universal recipe.

A few tips that help a lot:

• stay active, you do not learn much by only listening
• one hour a day beats the mythical two month break you will never get
• the best day to start was yesterday, the second best is today
• avoid random tutorials, choose sources you trust
• for your chosen expert area, invest in a proper course, it saves huge amounts of time and gives you the right mental model

todo: Find a way to make learning enjoyable. If it feels like a struggle every time, you will give up sooner or later. I became good at dataviz by exploring R package vignettes late at night 🤓. It worked only because it felt like a video game to me.
​
Btw if you want your speciality to be dataviz, my ggplot2 and matplotlib courses are very in depth & highly interactive! 😀
​
Stop scrolling. Stop watching TV. Go learning!

​

3️⃣ Apply your skills

Knowledge without practice does not move your career. Diplomas help far less than people expect. What convinces employers is proof that you can work with real data and solve real problems.

This is where side projects matter. You will not be paid for them, but they are the fastest path to credibility in my opinion. And they should all support your chosen speciality.

Some examples to spark ideas:
• I spent one hour per day for a few years building the R Graph Gallery and Data To Viz. Those side projects changed my career.
• Joseph Barbier, who worked with me for a year, built several Python dataviz libraries like pypalettes on the side. There are hundreds of R libraries that still do not exist in Python btw!
• I built a bot in R to harvest all tweets that have #surf in it, and built a map of where surfers travel.
• If you love dataviz, "competitions" are good opportunities too. Last year I won the Pacific challenge and it adds a nice line to my resume.
• I also tried to get rich by doing crypto arbitrage, but that one never worked. Not a problem, it shows I can build a bot and a clean report!

• Other: tidyTuesday? Kaggle? Exploring Our World in Data? Create PRs to open source projects?

​
There are so many ideas. If you think I should write about this to give a long list, please hit reply!
​
Whatever you create, polish it. Clean outputs, clear visuals, good explanations. Your future recruiter must think: "wow, they know what they're talking about".

One last thing. If you are switching from another field, look for elements in your current job that can support your next step. For example, if you have published scientific papers, turn part of that work into a clean data analysis report hosted on GitHub. It instantly becomes proof that you know how to analyze data.

todo: list or create at least 4 projects that illustrate your expertise

​

4️⃣ Show

Your projects only help if people can actually see them.
​
I would love to know exactly what recruiters value most, but here are a few reliable ways to build a strong online presence in my opinion.

LinkedIn​
This one is obvious. It is widely used, ranks well on Google and takes a few minutes to setup. So polish it. Keep your profile concise, make it easy for people to contact you, highlight your chosen speciality, and focus on what you can do for others instead of reciting your whole career. And the most important part: add testimonials.

Github​
If your future job involves code, you need a tidy Github profile. Put every line of code you write there and make it easy to understand with clear readmes. Github also lets you turn your data reports into public websites for free (example), which is a huge advantage. If you wonder how to setup Github to showcase your work, I explain everything here.

Others​
A blog is great for visibility. Researchers can use ResearchGate. Tableau users have their own gallery. Any place where your work can live publicly is useful.

todo: type your name on google. What is the first thing people see? Make sure it's memorable.
​
​

5️⃣ Search

You are now ready to start looking for a job, but this is the part where I have the least concrete advice. Two things come to mind though.

First, real human connections matter more than anything. Talk to people. Go to conferences and meetups. Reach out to someone who does work you admire and invite them for a coffee. It can feel scary or a bit cheesy, but most of my best opportunities came from people who knew me, or knew someone who knew me, in real life.

Second, do not quit your current job hoping you will find a new one quickly. It puts you in a tough position and often makes you less attractive to employers.

I wish I had more to offer for this final step, but this is honestly what helped me the most.

​

---

That's it.

To be honest, I never feel fully comfortable talking about non technical topics. I am much more at ease discussing good charts and code until late at night.

But I still hope this helps some of you get closer to your dream job.

Job seeking can be tough, so remember that you are doing great work and that the right data role will come. And if there is anything I can do to help, just hit reply. I read everything.

Time for me to sleep. See you next week!

Yan

PS: module 2 of ggplot2 Uncharted is finally ready. Sorry it took a bit longer than planned, but we are really proud of how it turned out. Hope you will love it. ❤️
​
PPS: I am also working hard on my portfolio tool. Any feedback, good or bad, would help me so much. Please let know!
​