So you’re thinking about attending a Data Science bootcamp? You’ve seen lists like this one saying that data scientist is the best job in America (and highly paid to boot) and you are constantly bombarded with ads about how Bootcamp ABC will train you and help you land your first data science offer.
It’s definitely a gamble to quit your current job, shell out somewhere between $15,000 to $20,000, go back to school for 3 months, and then job hunt for an undetermined amount of time post bootcamp.
Now that I have gone through the data science bootcamp experience and come out the other side, here is what I wish I had known about it and the ensuing job hunting process beforehand. Disclaimer: I attended Metis in San Francisco so if you choose another bootcamp and/or location, your experience may differ from mine.
You will learn a lot of cool things… but 70% of it will not be useful for landing your first job
Computer vision is super cool, so is text generation with an LSTM neural network. But there is a 99% chance that your interviewer will care much more about your SQL table joining skills than your knowledge and passion for deep learning.

I’m not saying that deep learning and artificial intelligence are not important to know (they definitely are), but rather that you will probably be interviewing mostly for data analyst roles (or data analyst roles disguised with the data scientist title) when you first finish your bootcamp.
And as far as I can tell, the key skills that you need for these types of roles are:
- SQL
- SQL
- SQL
- SQL (you need to be really good at SQL)
- Data cleaning with Python (NumPy and Pandas)
- A/B testing
- Data visualization (preferably with Tableau)
- Basic statistics (correlation, statistical inference, linear regression)
- Some knowledge of clustering and classification algorithms
Sadly, many of the things you spend significant time working on during your bootcamp (let’s be honest, training a neural net to quote Edgar Allen Poe is way more fun than running a hypothesis test of the difference between two means) will be under-appreciated by your interviewers.
And unfortunately, the one thing that you need to be an absolute ninja at, SQL, is barely covered and worked on during the bootcamp. Instead, you will most likely spend your first few weeks post bootcamp practicing SQL hard on your own. If you want to speed up this process, then devote some evenings during your bootcamp to learning and practicing SQL. My buddy Randy is working on a great Intro to SQL series here.
You will have to line up most of your interviews on your own
Your bootcamp will tout the transformative power of its alumni network, hiring partners, recruiting assistance, and interview prep.
Yes, there is definitely some career help but data science bootcamps are different from traditional software developer bootcamps – I am totally guessing but I would say for every 1 data scientist hired, there are probably 15 to 20 software engineers hired. As trendy and in demand as data scientists currently are, they still cannot compare to software engineers in terms of demand for the skillset – thus, you shouldn’t use your friends’ software engineering bootcamp recruiting outcome as your motivation to attend a data science bootcamp.
And the data scientist supply demand mismatch is not what it once was. Years of hype and high salaries have probably brought the market closer to equilibrium, at the same time that many companies are finding out that there are limits to what data science and machine learning can help them achieve.
So the moral of the story is that there are ever more of us fighting for a piece of a pie that is probably not growing as fast as we hope it is.
Personally, I went into the bootcamp expecting there to be a "Recruiting Day" or "Interview Day" near the end and I was disappointed. While we did have a Career Day where we presented our final projects to a few recruiters, I don’t believe anyone from my cohort got an onsite interview, much less a job out of it. And not for lack of trying – my classmates presented some really great projects and everyone schmoozed hard during the networking session. Unfortunately, many of the recruiters were either not hiring or trying to fill positions that required much more experience.

So what’s an aspiring data scientist to do? I have three suggestions:
- Spam that Easy Apply button on LinkedIn. LinkedIn has this amazing feature called Easy Apply that lets you apply for a job in literally 5 seconds (as long as your resume has been previously uploaded). Even with a really low response rate (and yes, you should expect a lower response rate for these listings because everyone applies since it’s so easy), you can still get one or two interviews per week. That’s because if you just devote 10 minutes a night to spamming Easy Applies, you will end up applying to 40 or 50 jobs each week.
- Lean heavily on your network if you have one. Grab lunch, have coffee, whatever you can with your friends and acquaintances that are either in the industry or at companies you are interested in. Yes, networking is no fun (for me at least), but it is a necessary part of the process. And tech companies hand out sweet, sweet referral bonuses these days so if you can demonstrate that you are a curious, humble, and reasonably knowledgable person, then people will generally be eager (and incentivized) to help.
- Don’t be overly stuck on that data scientist title. You shouldn’t expect your first job post bootcamp to be your dream job (but if you find it, then that’s awesome). Every new job is an opportunity to learn and build new skills and as long as you keep this compounding process going, you will end up in your dream role at some point. But I would highly suggest you to not be a job snob – rather, find as many avenues to gain analytical experience as you can.
Nobody will care about the projects you work on unless you make them care
One of the key selling points of the bootcamp was the chance to put together a portfolio of data science projects that I could then tout as evidence to companies that, "Hey I can do this data science thing!"
But looking back, it was naive of me to assume that just because I trained a model that produced good results, threw it on my GitHub, and wrote a bullet about it for my resume, that magically a recruiter or hiring manager would see it and be like, "Cool! We should give that guy a chance."
You need to market your projects and your data science knowledge. Personally, I do it through my blog. Here are a few Medium tips if you are an aspiring data science blogger:
- Publish on a publication like Towards Data Science, which has a large dedicated following. Much more people will see your work than if you just publish on your own, especially in the beginning.
- Use non-technical language, clear and simple examples, and enticing visuals to explain your work and ideas as most people on Medium are not practicing data scientists. This has an added benefit of making sure that you truly understand what you are writing about – learning to explain a complicated topic to a non-technical person is an incredibly effective way to build your own understanding and expertise. And you will get more reads too.
- Don’t just regurgitate your project step by step. This is not a high school science project report. You are writing for an audience that wants to be both entertained and educated. If you want to highlight your project work, you might consider how a certain aspect of your project demonstrates a key data science principle. For example, you could write about your linear regression project while at the same time teaching your readers why it is critical that your model upholds the assumptions of ordinary least squares. Writing this way generally makes for a much more interesting read than "First I did A, then I did B, then I did C, etc." And more interesting means more reads and ultimately more exposure for both you and your work.
However you market your projects and yourself, just remember that unless you make them care (by presenting your process and insights in an eye-catching manner), they won’t care about all your hard work.
You need to know the business (of every place you interview at)
This is harder in the Bay Area (versus somewhere like Houston where every company is in the oil and gas sector). There are so many startups out there, each attempting to disrupt a particular industry with its own unique business economics.
So unfortunately if you want to work in tech (especially tech startups) in the Bay Area, you will end up doing a lot of reading and studying up on the competitive dynamics of various industries. Wikipedia and Crunchbase are good places to start. Business strategy blogs like the a16z blog and Stratechery are great for deeper dives into industries you are interested in or interviewing for. Morningstar also does a surprisingly good job of analyzing companies and industries – pay special attention to their discussions on what drives a firm’s economic moat – the characteristics that allow a company to sustainably earn a high return on its capital over time.
The first question I was asked in multiple interviews was, "So what do you know about us?"
Generally recruiters and hiring managers expected me to know at the bare minimum the following things:
- What business is the company in, and what products or services does it sell?
- How does it make money? What special features of the firm’s business model allow it to separate itself from competition and earn profits (if no profits, then what allows it to grow faster than its competition)?
- Knowledge of how data science is currently being applied by the company to gain a competitive edge or overcome business obstacles.

Job hunting is its own full time job – at the same time if you don’t work at it, your hard-earned data science and coding skills will atrophy
A lot of data science bootcamp grads are career changers, with many coming from non or at least less technical backgrounds. Gaining traction in the job search will take a lot of time and effort.
At least for me, job hunting with all that it entails (applications, cover letters, coffee meetings, phone calls, interview prep, the actual interview itself, and thank you notes) was extremely time consuming. Then on top of that, I also needed to maintain my blog.
Before I knew it, more than a week had passed with me barely touching Python. At that point, I forced myself to step back from the job hunt and work on an old data science project.
As much as we want to land that job, it’s important to remember that we are not data scientists because some company was willing to pay us to be one – rather we are data scientists because we practice data science frequently (coding up new projects, participating in Kaggle competitions, researching and reading about new algorithms, etc.).
So keep up that Python, keep working on those projects, and keep showing your passion and eventually someone will take a chance on you. Cheers and good luck!
More Data Science and Analytics Related Posts: