OK OK, I'll admit it. I'm on a contrarian streak. For good reason - I want to help you with your analytics career and there are common potholes such as overrated technical ability. Analytics degrees are a close second and worth an in-depth discussion.
When I mention "degree" I mean any of the following:
- Bachelors/Masters in Analytics
- Analytics Boot Camps
- Technical Certifications
*There are a few exceptions to this advice, though they are very case-by-case. There may be a specific position you want at your company that requires a degree to get in or you may have a personal to accomplish. I'm not speaking into those situations but still want to acknowledge they exist.
The Allure of Education
It's logical why many have a thought process like this:
- I am interested in analytics
- I do not have analytics experience
- Hiring managers want to see experience and/or education
- Education is the next best option
- I will fill in gaps in my resume with education
At face value, this makes complete sense. In other career tracks, education teaches crucial skills and gives you an entry into that industry. Want to get into law? Get a law degree. Want to become a doctor? Get a medical degree.
This is absolutely not the case in analytics. A Masters Degree, Analytics Boot Camp or MSSQL Certification will not give you a leg up for analytics positions. I see post after post after post on data science forums discussing analytics education. A key assumption is rarely called out: "Education will help you get an analytics job."
Why Classes Struggle to Teach Analytics Skills
I had the privilege representing BI/Analytics on a panel for the University of Washington Information School. I centered on one basic point: it is near impossible for a classroom setting to prepare you for the reality of an analytics career.
Think of it this way: in college, the "game" is well-known. The teacher gives you specific concepts. Your job is to apply those concepts on your homework, tests and/or projects. The requirements are clear and tie back to the class syllabus. Data is typically clean or requires trivial amounts of cleaning to get ready.
Analytics careers are nothing like that. I wrote about how ambiguous data problems are. There's no syllabus. Clear questions are rare. Even if questions are clear, your stakeholder often asks the wrong question. Data may not exist and any existing data is a mess. The world is ambiguous and cloudy and hard to navigate.
Imagine a college class that tried to replicate this. No syllabus. Little to no data provided. You may or may not have a test, and that test may require you to answer questions not even on the test. Even if there were questions, they may not be the ones the teacher wants you to answer. What a mess of a class!
I'm not sure how to structure a college course to capture the ambiguity in the every day life of an analyst. As Jacob wrote, there are four key soft skills for analysts and I'd be interested to hear of any creative strategies from teachers/professors to teach them. Certainly some get closer than others, but no matter what there is no replacement for the real world.
Why Degrees Don't Matter
You may have already connected the dots. If courses can't teach key analytics skills, then various degrees will not make a resume stand out. It's rare for technical ability to stand out as the reason to hire someone.
Combined with the time & expensive involved with degrees, their value diminishes. Put another way, if you can get better experience AND get paid for it, consider that option first.
In Conclusion - What Now?
Experience is king, period. You may be asking "But how do I get experience without getting my first job?" Great question! This is what I referred to as the 'Great Filter' on landing your first analytics job. That post will cover most of what you should do instead of getting a degree.
A note from Jacob: For more on this - lots of good discussion on data twitter & in the Locally Optimistic slack. A snippet of a thread just yesterday is below.
Was just talking to someone looking for tips on preparing for data science interviews and realized I couldn't give them any concrete answers ("should I study stats? programing? analysis? which models?") since every single interview is radically different. Unless you're preparing for a FAANG-style interview where they literally give you a packet of possible questions and guidance, I have no idea how any of us know what to study and get jobs in this industry. I was reminded of @tdhopper's great post on this topic. https://tdhopper.com/blog/some-reflections-on-being-turned-down-for-a-lot-of-data-science-jobs
Originally tweeted by Vicki Boykis (@vboykis) on November 9, 2020.