home > Research&Insights > The Data Science Gap

The Data Science Gap

Every other day it seems there is a new article about how data science is the best field for job prospects.

Both in demand and well paid, it looks ideal for both students on the hunt for job security and workers seeking better pay. Advice to acquire data science skills hasn’t been ignored — across the world, thousands of students are now enrolled onto analytics programs at universities and online.

Despite this, attention has been drawn to a gap between the number of jobs on the market and the available candidates for them. Should graduates be worried about this data science gap, and what can they do to improve their chances of getting a job?


Yes, the demand for data scientists is still high…


There is some misinformation going around that data science has had its heyday. Roles in the field have “grown over 650% since 2012”. What’s more, LinkedIn’s 2017 Emerging Jobs Report on the fastest-growing jobs in the US (based on their user data) highlights that “[t]ech is king,” and the platform states that jobs with titles such as machine learning engineer, data scientist, and big data engineer are highly sought-after across a range of industries.

This report, which analysed data from job posts on LinkedIn over a five-year period, notes that 6.5 times as many data scientist positions are posted today than in 2012. Even better, of their 10 listed “most common skills among emerging jobs”, 3 relate directly to data science (Python, software development, analytics).

…but there aren’t enough people with the right experience


Three years ago, Jim Davis, Executive VP and CMO at SAS said: “If you want to get a job quickly, figure out how to become a data scientist” (Source). Or go to the right place to up-skill yourself — like DataScienceGO2018.


The problem is that the field has grown too fast: there are now far too many data scientists with very little experience heading into a job market that has very few experts, leaving us with a ‘bottom-heavy’ candidate pool. This means that, while there are still opportunities in data science, individuals looking to make the most of the gap need to be clever about carving out a career path and getting ahead.


…and there’s a lot more competition for the best jobs


Times have changed. When the term was still relatively new, companies were accepting candidates with only a basic knowledge of data, and getting them to learn on the job. Now companies often won’t hire data scientists unless they have a much deeper knowledge, say, of coding and statistics. Demand is still high, but now the bar is even higher. Says Kevin Safford, Sr. Director of Engineering for Umbel:

“Every year, PhD graduates in statistics, econometrics, hard sciences, and computer science — many focusing specifically on machine learning — discover they have zero interest in academia and enter the workforce.” (Source)
This means that candidates now have to face an incredibly competitive market. An application that may have looked attractive five years ago might today be put on the rejection pile.

BUT… is this really the full picture?


The Huffington post claims that there are approximately 1.5–3 million data scientists in the world right now — are there really no candidates with the right experience?

What if I told you, the people responsible for this “Gap” aren’t really the data scientists; rather the directors, HR departments and even the recruiters in different companies.

Why?

The truth is, data science has become a buzzword, a hype. For 5 years now, it has been deemed “the sexiest job in the market” by Harvard Business Review and companies now rush to include a data specialist in their ranks. But the real problem is that they don’t know what a data scientist really is, what they do, how a team is configured, what makes a great data scientist really “great” — there is a huge gap in knowledge of the profession itself.

There are plenty of data scientists out there, there are thousands of talented professionals that could easily flip a pre-data business up on its feet, but they are rarely given the chance to prove themselves. Most companies, in a rush to “hire a data expert”, assume that they need a 5–8 year experienced individual to solve all their problems (even though, ironically, the field hasn’t been alive for many years) and they conclude that every data scientist out there with less than that experience is no good.

Fortunately, this means opportunity


It wouldn’t be fair to push blame and say that everything is due to ignorance on the profession — we data scientists are also to blame. 
 
 It shouldn’t be surprising that HR has probably been doing its job in a traditional way, therefore it’s completely normal for them struggle with these new job positions. They are looking for degrees in our data science careers that don’t yet exist, they want experience where there was no field to get it from, they want skills that they don’t clearly understand… yet.

It’s up to us to showcase our strengths, capabilities and what we can really bring to the table. Data is the future of business, there’s no going around it — and we know this, so let’s show everybody else what we are made of.

But how do we do this?

We should try to narrow the data science gap.

6 ways to fill the data science gap

So how should we position ourselves to ensure that we get the job we want? Should we become great at visuals? Should we acquire a comprehensive knowledge of analytics software? Drop the crystal ball for now and consider your employer: what all companies want from their data scientists is the ability to solve practical problems, and better yet: to be able to communicate what they have found. If we can use data to answer real business questions, then we will have a far better chance of getting the job we want.

But for that we need experience and knowhow; there are no [longer any] shortcuts to becoming an in-demand data scientist. Here are six approaches we can take to improve our chances:

1. Understand the field


Data scientists are needed across the board, from pharmaceuticals to sports. A sales company might want to know how they can tailor their marketing campaigns to target the right customer segment. A financial conglomerate might want to use their historical data to help them reduce risk. A videogames publishing house might want to know the steps they can take to increase the number of loyal players. A governmental institute might want to see how they can start to implement smart technology into their city.

The better you understand a field, the more likely you’ll succeed in managing its data. There is no point in taking on a data science role in a bank if you don’t understand how the finance industry works. You must take care to learn about its practices and methodologies in order to answer those all-important practical business issues.

2. Take a course


You need to be able to be quick, in a field that moves as quickly as Data Science does. Tomorrow, you’ll find that some of the topics that where important today, no longer are; so the best way to keep being relevant in the field, is by keeping that hunger for knowledge alive and up-skilling yourself. 
 
 Morning Future suggests that the best way to prepare is by taking a targeted course because “that’s where four out of ten companies are going to look for the data analysts of tomorrow.”

Whether online or on campus, data science courses with a good reputation are the best way to get at least a primary understanding of working with data. Many online courses come with certificates of study to prove you have paid attention and done the coursework — they might not be the most appreciated at the moment (trust me, I know) but they will be in the future.

3. Get a mentor

Mentorships are great ways to seek advice and clarity on job prospects and careers. Many data scientists are happy to develop promising new candidates and, in my experience, they welcome the idea of giving advice to others. We are a group that thrives on sharing information and best practices. It’s after all the best way to improve!

4. Read the news


If you haven’t heard that data = speed, then you’re one step behind.

Data moves fast and hundreds of people are working on thousands of projects — and who knows? maybe that missing piece that you’ve been struggling with that solve that issue that you’ve been having, might have been solved this morning by someone else!

As a field that’s been built upon a premise of collaboration between its members, data science constantly benefits from the work of others allowing each person to build grander and more impactful things.

Keeping up with the news, subscribing to the right blogpost, listening to the perfect podcast or receiving the right newsletter might be the solution you might be overlooking.

5. Apply to the right companies


Data science has attracted a lot of press for being a fix-all for companies and their increasing amount of collected information, but this is a double-edged sword — most companies ultimately still don’t know what they are looking for in a candidate.

If you want to protect yourself against working for companies that don’t understand what they want, Robert Chang recommends steering clear of meandering job specs. Sage advice: prospective jobs that cite a long list of methods and software you need to have knowledge of only shows how little a company understands of its data strategy, its data requirements and not to mention “they’ll hire anyone because they think that hiring any data person will fix all of their data problems.”

6. Network

Don’t like meeting new people? Reconsider if you really want to be a data scientist. Remember point 3 above: data science is ultimately a social field. Networking doesn’t have to be face-to-face; you can start from the comfort of your own living room. Speak to other data scientists on Twitter, reach out to them on LinkedIn, or join a discussion on Reddit.

If you want to step up your game, there is a super friendly, informative and relaxed event coming up that can get you started in building your networking skills: DataScienceGO, a career-focused data science conference.

More than ever, new data scientists must learn how they can create a reliable pathway to achieving their career goals. This is exactly what we had in mind while devising the individual elements of DSGO: career-centric approaches to data science. DSGO connects participants with the clients, mentors and partners suited to their skill level, experience and interests. The event will host technical discussions, renowned industry speakers, hands-on activities and motivational sessions to get participants started in data science and give them the inspiration for moving forward in the field.

For a current list of confirmed speakers for DataScienceGO 2018, visit our conference website. Subscribe to our newsletter to keep updated with new speakers and scheduled events. And hurry, our early bird tickets are selling out fast!

Don’t quit now!


Ultimately, as good data scientists, we must look at the evidence rather than focus on conjecture.

The truth is that data gives companies the information they need to get ahead of the competition, and they therefore address the single most important need of business: to make money. We are not using less technology than we were five years ago. Smart devices, loyalty cards, social networking platforms, internet searches, all of these things are only going to generate more data, year on year. People will always be needed to manage an increasing rate of information. The evidence shows us that jobs in data science are still set to grow, and that the field is relatively safe from automation.

Yes, the field has become much more competitive, but don’t let the competition beat you before you’ve even attempted to get anywhere. There are plenty of jobs going, and with a bit of effort, hard work and the right attitude you are still highly likely to succeed, even without the work experience.



By Kirill Eremenko