Chapter 2: Why I chose to be a self-taught data analyst

I chose to be self-taught, but it’s certainly not for everyone. It’s something I felt suited my style. I’m an independent and avid self-learner. I did research on materials prior to taking the leap. I had enough savings to last me comfortably for a few years. This gave me a comfortable cushion to purely focus on learning. I knew in the worst case I could go back to my old role.

I’m not going to lie. Becoming a self-taught data analyst is hard. But it’s definitely possible to make it work, now more so than in any other time in history. The barriers to entry have never been lower. The challenge here is NOT access to information. It really is how to assimilate the right information. These were my considerations before I took the leap.

Why I chose the self-taught route, and is it for you?

The sole reason I chose the self-taught route is that it’s doable. I found the trade-offs and risks lower than other options. Let me explain.

Information is free

The best learning materials are available online, and they’re mostly free. Harvard’s CS50 intro to computer science, Andrew Ng’s detailed machine learning course, MIT’s course on applied probability, An Introduction to Statistical Learning. These are some of the most comprehensive starter materials on math and computer science, and you pay nothing for access.

Back in university, I’ve had some poor professors where going to class was akin to throwing time away. Their explanations were poor and they were bad communicators. Bad professors don’t know the material good enough to explain it simply. Here’s an except from a 1996 issue of Caltech’s Engineering and Since magazine about Richard Feynman, one of the most renowned physicists:

“Feynman was a truly great teacher. He prided himself on being able to devise ways to explain even the most profound ideas to beginning students. Once, I said to him, “Dick, explain to me, so that I can understand it, why spin one-half particles obey Fermi-Dirac statistics.” Sizing up his audience perfectly, Feynman said, “I’ll prepare a freshman lecture on it.” But he came back a few days later to say, “I couldn’t do it. I couldn’t reduce it to the freshman level. That means we don’t really understand it.”

I’d much rather take a good online course or book any day. Why? I’m guaranteed a good explanation. It helps me learn quicker. I’m more interested in learning meaningful stuff than getting a cert. Being a data analyst is about adding value and not being a good salesmen. You don’t need a fancy degree if you want to add value.

Low risk

Being self-taught is low risk. What’s the worst that could happen? I spend a few months to acquire new skills. If I realize I don’t like the work or if I fail to learn and think that it’s too hard, I’ll just go back to my old job. I like the optionality.

Self-taught can be faster

Being self-taught allows for focus. University curriculums are broad - they touch on many different areas, from “Data Analytics in Business” to “Deterministic Optimization”. Such courses may help you some day. But it’s akin to spending your weekends learning Polish, on the off chance that you’ll visit Poland some day. It’s useful, but of low practical utility. Being self-taught allows you to build up key skills that companies are looking for faster than you would in formal education. It also allows you to go deeper by doing more of the same type of projects if you choose.

More than anything, I knew it was possible. In fact, numerous notable contributions in history were made by autodidacts. Many famous names in science, music, and artists were self-taught.

Side benefit: Acquire life skills

I’ll wager that by choosing the self-taught route, you’ll pick up much more than just data skills. It’s likely that you’ll acquire the meta skill of learning how to learn, as well as build a level of self-confidence that allows you to pivot wherever you want in future. Doing a hard thing once builds upon itself. This meta skill and confidence will compound over your life.

Other options

I can’t speak for the two options below as I never went through them, but this was my rationale.

University

The opportunity cost is high in both university fees and time. Also, I’ll be locked in a program, something which didn’t appeal to me. By choosing the self-taught route, I really was betting that a cert would not mean as much compared to actual experience.

But if you can get into a top school, or if you get subsidized, or you just want a long break - it might be worth it. You’ll get a network in university too which is valuable.

Bootcamps

Bootcamps provide a set curriculim and a peer group to study with. Attending one basically guarantees you a final “capstone” project and some python skills. It’s low risk.

But I knew that bootcamps would not provide much more than what I could get for free. I did not choose bootcamps as I knew I could work alone, and I didn’t like the idea of going through a set curriculum. I did extensive stalking on bootcamp portfolio websites and invariably found projects that were similar. Recommendation engines were almost the default capstone project. I’d rather learn what I wanted and do it at my own pace.

With that said, bootcamps have a network of companies to place graduates. That might help if you want hand-holding throughout your journey.

It’s your choice

It’s ultimately your choice. You have to take a look at who you are, and what you’re comfortable with.

If you’re still considering the self-taught route, read on.