Opinion
There’s a chance that the supposedly hottest job of the 21st century might not be the right one for you. I’m not saying that learning about Data Science isn’t worth it (just the opposite), but it’s better to get involved in the field if you actually have some problem you’re trying to solve.

Note from the author: This is an opinion piece, so it’s probably biased to a degree. Jobs in your country and with your skillset may vary. We don’t see the world through the same eyes. Please leave your thoughts and experiences in the comment section.
After being in the industry for a while now, spending 10–12 hours a day exploring the field, and writing close to 100 articles about data science, I feel qualified enough to speak about things that were quite disappointing but also never mentioned during my studies.
For me, practical data science is one of the most interesting fields – especially with the advancements in deep learning.
So what? What’s the point of doing data science just for the sake of data science? Well, if you’re not a researcher, there’s almost no point in being in the field full time.
What’s the reason you’re spending hours and hours training a model? Is it because you’re planning to deploy it in a web and/or mobile app? Is it going to solve some real problems? Or will it just sit idle in the notebook?
For a while now I’ve been feeling like data science is becoming yet another branch of software development. And that’s not necessarily a bad thing, because model training and optimization are nowadays pretty much automated (with access to the right tools), and the only task you have to do as a "data scientist" is to prepare the data in the right way and also to present your work to superiors.
And for me, that’s great because I can spend more time doing interesting stuff – like app development – hence delivering the whole package and actually solving problems, and not leaving models sitting idle on my laptop.
But if you plan to work from 9–5 for someone else, then make sure to read this article, as it will speak about some aspects which aren’t so peachy. Also, as stated in one of my previous articles, you should know what are the benefits of enrolling in the field and want to learn about the possible drawbacks.
Okay, so without much ado, let’s get to the reason number one.
1 – You don’t care about the business
Let me tell you a story. Besides blogging here at Medium and TDS, I also have a full-time job as a data scientist. The company I work in isn’t specialized in any branch of the industry, it’s just an IT company primarily focused on consulting.

What does that mean for me? That means that I don’t work on one project, or one product, but instead work on 3–5 different projects from different industries. And as you’d expect, I’m not a domain expert for any of those.
And that’s where you might end up disliking your data science job – as you don’t know much about the domain you’re working in, but also aren’t eager to learn about it because it bores the hell out of you.
Now, this might not be the case for someone working as a data scientist in a domain of interest. Let’s say you’ve got your education in the field X, but also know math and stats pretty well, so you were hired in your domain to do more data-heavy work. That’s entirely different.
Here I’m talking about generalist data scientists, the ones jumping from project to project never feeling like they are actually contributing.
2 – You don’t see the big picture
Unlike the previous reason, this one applies to both generalists and specialists in the realm of data science.

Let’s say you got invited to a meeting from a project team because their app could benefit from your "data science magic". Questions like these are inevitable: "can your model do this?", "can you implement these modern neural networks on our 10KB dataset?". Maybe not the last one in the exact sense, but don’t be surprised if someone needs a predictive model and has only 30 rows of data.
And that’s the problem because somehow you should deliver a state of the art solution in no time with knowing absolutely nothing about the project and its structure – hence the big picture problem. Good luck with that.
Things here are once again better for a specialist data scientist because having at least some domain expertise and knowing how you fit in the whole picture should make it easier for you to do the job right.
3 – Your boss is a businessman
Or businesswoman.
The point is – your superior is not an expert in any data-related field and knows about data science as much as your average Joe can learn from the news.

Note from the author: The amount of impact this reason has can be different from country to country, as work and life culture may differ severely.
And this poses a potential problem because your team might be given a task which implies months worth of research, but according to the business that just isn’t an option.
Hopefully, you’ll have a department manager who understands how much time and work is required for a particular solution, and he/she will be able to talk some sense into his superiors, but you can’t count on that assumption.
And that’s where having great people skills comes in handy. Explaining why a month or two was spent on finding 100 things that don’t work isn’t a pleasant thing to do, but having those people and presentation skills might just save the day.
Conclusion
First of all, this was an opinion piece. Make sure to leave your thoughts in the comment section below.
In my opinion, knowing a lot about data science and predictive modeling should make you a viable employee, but only if there’s an actual data science-related problem to solve.
At the end of the day, data science is just a skillset, worthless if not applied to a business problem. If you don’t care about that business, or just don’t see how you fit in, things probably won’t end that well.
Thanks for reading.
Loved the article? Become a Medium member to continue learning without limits. I’ll receive a portion of your membership fee if you use the following link, with no extra cost to you.