Becoming a Data Scientist: Master’s Program, Bootcamp, or MOOCs?

This post is contributed by Irmak Sirer, a partner and data scientist at Datascope, where he solves business problems with data by designing analyses and interfaces. Irmak has helped companies across industries solve problems with data, from small companies to members of the Fortune 50. Working with Metis, Irmak is an instructor for their Data Science Bootcamp program which will be starting on September 2nd in New York City. You can read the full version of this post on the Datascope website.

One of the most frequent questions we hear, right behind “so, what exactly is a data scientist” or “what makes a great data scientist”, is “how do I become one? I should probably just get a Master’s, right?” Perhaps not anymore; rising costs, changing demand, and the Internet are disrupting this traditional path and providing two viable alternatives. At one extreme, self-learning through Massive Open Online Courses (MOOCs) give access to courses at an extremely low cost (often free), but leave it “as an exercise for the reader” to identify a suitable set of courses and tools to round out a coherent skillset. Bootcamps offer a middle ground where students can pay for a structured learning environment at a far more affordable rate compared with obtaining a Master’s Degree. So, “which path do I take?”

We think the answer to that question largely depends on the student. In some cases a student will prefer attending a bootcamp whereas in other cases a student will prefer receiving a Master’s at a university or taking university courses online through MOOCs.

Here at Datascope we see great benefits from the bootcamp format, so when Metis (a part of Kaplan) contacted us about partnering to design a data science bootcamp, we jumped at the opportunity. We thought we could take all these points we see as the advantages of the format, and elevate them as much as we could. So, we designed a course that would give aspiring data scientists a lot of experience with 4-5 projects, and a guided route of several core data science concepts and approaches. Participants can quickly build the necessary foundation without the burden of teaching herself everything or paying the handsome price of a Master’s program before realizing her dream job. If you’re interested, our Data Science Bootcamp program is starting on September 2 in New York (applications due by August 11), and you can learn more about it here.

Since there are many things to consider when choosing which program works best for you, in a separate post, we do a thought experiment to compare the three experiences for a fictitious aspiring data scientist named Audrey. For the sake of brevity, the following table summarizes our thinking about what each of these experiences is like and, more importantly, who they are ideally suited for.

	Masters	Self-taught (MOOCs)	Bootcamp
Learning	Theory-rich learning	Self-guided learning	Experiential learning
Teachers	Live university faculty professors	Recorded university faculty professors	Practicing data scientists
Outcome	Diploma	Certificate	Portfolio of projects
Duration	9 - 20 months	6 - 18 months (part-time)	2 - 3 months
Tuition	$20,000 - $70,000	$0 - $500	$0 - $14,000
Networking	1.5 years of social networking	Isolated; no in-person networking	Collaborative networking
Projects	Internship + practicum projects	Projects on own time	Projects built in to experience
Job hunt	University-wide recruiting day	Self-driven job search	Hiring day organized by bootcamp; talent placement manager helps with hunt
Ideal for	People that enjoy immersing themselves in campus life and want to take time to let the new material absorb while learning in a structured environment with the full credentials of a University degree.	People that thrive with ambiguity and self-guided environments and are motivated enough to design their own curriculum around their own strengths and weaknesses.	People that want to switch or accelerate careers ASAP and want to have confidence that the switch will result in a job they will like while learning in a structured environment.

As technology increases the rate of change of society, the most successful workers will be those that can quickly shift to new specialties and learn on the job to meet market demands. In our opinion, the bootcamp format provides the benefits of personalization, credentialing, and social learning that a Master’s degree offers, but at an accelerated rate with experiential learning. Sure it is more expensive than being self-taught, but the connection with employers and the guided, experiential learning process increases your confidence to tackle the uncertain prospect of making a career switch.

To become a data scientist, you don’t need to have postgraduate degrees, or 20 years experience, or be proficient with every data-related technique and tool under the sun. What you need is to have enough baseline knowledge and experience, and the skill to constantly adapt and learn. Bootcamps, in our opinion, are the perfect medium for making the transition.

Tell Your Kids to be Data Scientists – Not Doctors

Recently I had the pleasure of being interviewed by John Phillips at CNBC about our data scientist salary study. His article, Why Your Kids Will Want to be Data Scientists, was published at the end of May, and in it he raised a very interesting point:

“According to Burtch Works’ 2014 study of salaries for data scientists… those responsible for a team of 1-3 earn [a median salary of] $140,000 and those responsible for a team of 10 or more earn $232,500.

By contrast, the mean average annual income for a lawyer in America was $131,990 in 2013, while doctors earned $183,940, according to data from the U.S. Bureau of Labor Statistics.”

Did you hear that? Data scientists earning more than doctors! For complete salary information for data scientists, Big Data professionals and market research professionals, download the full reports for free here. Salary is not the only reason however, that I would recommend encouraging your children to pursue statistics and coding over going to medical school.

How Data Scientists Are Supplementing Doctors

There are big changes happening in healthcare right now, and the implementation of EHR (electronic health records) in particular is a great example of how data scientists will be working with doctors in the future. The move to EHR is picking up steam, and the Center for Disease Control reports that 78% of office-based doctors are using EHR as of 2013, with that number only expected to grow as practices will face penalties for non-compliance. All of these electronic patient records spell out Big Data for the healthcare fields, and data scientists - like all quantitative folks - love data. These medical data could not only offer tremendous insights that change the face of modern medicine, but also offer rewarding opportunities to the data scientists who must decipher the data.

Patient care also stands to receive enormous benefits from data science. Venture capitalist Vinod Khosla was recently quoted at the Standford University School of Medicine’s Big Data in Biomedicine conference saying, “Humans are not good when 500 variables affect a disease. We are guided too much by opinions, not by statistical science.” While a doctor may be trained to look for many factors when diagnosing an ailment, some of these diseases are impossibly complex, and patients could stand to gain faster, safer treatment if left in the hands of a well-developed machine, or even a physician aided by one. For example, IBM’s Watson is already collaborating with Memorial Sloan Kettering Cancer Center to help doctors make better cancer treatment choices. The human interaction between patient and physician will continue to be important, but data scientists will have a measurable impact on the future of healthcare.

Career of the Future

One of my predictions for the analytics hiring market this year was that data scientists would be embedded in analytics groups, and with the internet of things, the increase in wearables, social media sentiment analysis and many more applications for data science, it’s no wonder this career has so much buzz around it. With the increase in demand, shortage of talent, high salaries and applications in every industry, data science is becoming a good option for career success. The road to that success begins with a strong early foundation in math, and (perhaps) some nudging from the parents. I can’t tell you how excited I was when my daughter, Becky (who started college this year at Macalester), changed her major to math, as I believe strongly that this is a career path that offers a bright future. Who knows, maybe one day I’ll have a data scientist in the family!

Tips for Hiring Data Scientists

This post is contributed by Frank Lo, an experienced data science professional and friend of Burtch Works. Frank is currently the Head of Data Science at Wayfair, as well as the founder of DataJobs.com.

 There is so much hoopla around the need to hire data scientists -- but amid all the frenzy, I notice a major disjoint between what companies think they need and what they actually need to leverage data science for business value. I come from the background of leading a data science team; on top of cultivating the team and diving into the nitty-gritty of data, I spend a lot of time recruiting, trying to find new rock-star talent. I'd like to share a few things I've learned along the way around what to look for.

  Look for quant experts with business hustle.

 First, let's think about what is data science. It is not only a combination of technical and quantitative disciplines, but also the acumen to leverage STEM skills to transform business. Too often, we focus too much on pure engineering/math ability, kicking business smarts to the wayside. On the contrary, I think business acumen is one of the most important traits of effective data scientists -- so much so that I filter out strong tech/math candidates if they have trouble thinking through the business applications of their quantitative work. Ultimately data scientists create value by being consultants to the business. Data mining and predictive analytics by themselves are not the point, but rather the means to enable intelligent strategy development. Look for people who are good at all of the above, including business.

  The #1 intangible is intellectual curiosity.

 The spirit of data science is discovery. Given a mountain of data, what inferences can we make? What truth is revealed or predicted? The strongest data scientists are motivated by this curiosity to explore data in very creative ways. When I recruit for my own team, I look for people who are not only good at answering questions, but who want to ask their own questions. This genuine inquisitiveness is rocket fuel in driving a data scientist's search for meaningful discoveries in data. It is so critical to the role that we turn away candidates who cannot demonstrate that they are brimming with intellectual curiosity.

 How to screen for this intangible? One interview question I ask everyone: "tell me about a data science project or investigation that you initiated on your own, outside of school and work." I look for candidates who can respond to this at length, diving into quant, tech, or business problems they were so intrigued by that they carved out their own time to pursue. It is a great signal they will thrive in a data science role.

  Don't filter candidates based on degree.

There is a notion out there that the best data scientists are the ones with Ph.D's. From my experience interviewing and evaluating candidates, it is my opinion that academic degree is the last consideration that matters. Some Ph.D candidates have very well-rounded skills and do become top performers. Though, I've found that many others find themselves mentally stuck far down the academic rabbit hole, and have difficulty translating their focused depth into value in a business environment. My point is: degree by itself is a very incomplete indicator, so don't filter candidates on it. Consider all academic backgrounds, zeroing in on well-rounded skills paired? with the right intangibles (i.e. intellectual curiosity, business acuity, etc). You'll find that many star data scientists are self-taught hackers and mathematicians, who may not have wanted to sacrifice work experience for academic credentials, but have a very complete data science knowledge base.

 Clearly, there is a lot to consider in how to hire solid data science talent. The reality is that data science is very multidisciplinary; you're looking for a blend of skills that can reveal itself across a wide range of academic backgrounds and professional experiences. But as long as you have a solid understanding for the nuance in what makes a good data scientist, you'll have an easier time trying to identify the right people for your team.

Keep an eye out for our Data Scientist Salary Study - a follow up to the landmark Burtch Works Study: Salaries of Big Data Professionals - which should be released in the next month.

My Predictions for the 2014 Analytics Hiring Market

Thanks in part to the media coverage of Big Data, analytics has become inescapable. To those who have worked in and around this industry for years this information is nothing new - merely more visible. Every day I’m hearing about more and more businesses warming to the idea that there are actionable insights readily available in their data if they know how to use it.

It is an exciting time to be a statistician, data scientist, or analytics professional, and if you thought 2013 was big, 2014 promises to be even bigger. Without further ado, here are my top predictions for the analytics hiring market:

1.)  Real-time analytics becomes increasingly important to business intelligence – Industries like the gaming industry already give their customers incentives in real-time and retailers are quickly moving toward real-time couponing. Those of you with real-time data experience will be in high-demand for these growing roles.

2.) Data scientists will be embedded in analytics groups – In addition to analysts who can build complex models, there will be an increasing need for data scientists, who with their background in analytics and computer science can wrangle massive, unstructured data sets. Companies hoping to extract the full benefit from their data sets will need a combination of both groups.

3.) If you don’t have digital you will be a dinosaur! - If your skill set doesn’t already include digital analytics your career plan should be to get experience and soon. Without it your skills will become outdated and obsolete.

4.) Sentiment analysis will continue to erode traditional survey methods – Use of social media data and other “free sources” has the opportunity to overtake traditional survey methods, which will impact the marketing research talent landscape.

5.)  Corporations will bring on internal staff to replace high paid consultants- Consultants got the ball rolling, and now companies will seek to take on full time staff.

6.) However, this will not adversely affect the consulting industry – Demand is so high for analytics experts (and continuing to grow) that there will still be companies looking to small and large consulting firms. Despite some companies hiring full time staff, the consulting industry will continue to experience growth. Heavy travel schedules will continue to be the norm for these opportunities.

7.) The trend towards wearables and PEDs will create new positions- If the buzz from CES2014 is to be believed, the wearables trend is increasing. As with any new technology that creates massive amounts of data, analytics positions will be created in the industry to make sense of and capitalize on all that information.

8.) Startups’ grip on the hiring market will loosen- I have spoken with candidates who have worked with many failed startups, and I think that this year the tendency towards hopping from failed startup to failed startup will begin to shake out. Candidates who have been through the failed startup cycle will begin to realize that to build a career it will be important to learn from an established group.

9.) Graduates from the new analytics programs will begin to hit the job market - This is not so much a prediction as it is inevitable, and I am very interested to see how well these new programs have prepared their students to enter the quantitative workforce.

What trends do you think we will see in the analytics hiring market? Let me know what you think in the comments below.

 

 

Burtch Works' Most Popular Social Media Posts from 2013

2013 was a busy year for Burtch Works’ social media accounts, and I wanted to take the time to revisit some of our most popular links this year. 

Our top links include blog posts on topics from resume writing to data science wannabes, as well as links to our original research, including salary studies and a flash survey of our network. 

Here is our social media year in review:

1.) The consulting trend is really gathering steam but the opportunity is not for everyone, which I addressed in Should You Take That Consulting Role? Here's Why or Why Not.

2.) With networking being more important than ever and holiday parties around the corner these 18 easy conversation starters from Careerealism that I posted were perfectly timed.

3.) Apparently everyone is thinking about retooling their resume, because my latest blog post Need to Rewrite Your Resume? Four Tips Before You Submit is one of the most popular posts on our social media this year.

4.) Are you a real data scientist? Or just a Data Wannabe?

5.) McKinsey put out an informative infographic on Big Data and ROI Big Data Big Profits.

6.) Lou Adler once again published a great article on LinkedIn with 5 Things You Must Not Do In an Interview, and 5 Things You Must.

7.) The Burtch Works Study: Salaries of Big Data Professionals was released in July, and Silicon Angle ran a story about lucrative salaries for foreign-born analysts.

8.) Our Marketing Research Salary Study was released in October, both salary studies and their webinar presentations are available here.

9.) Back in April I addressed the lack of urgency to hire in Help Wanted But Hiring Slow.

10.) While catching up on TEDxTalks in September I watched Big Data, Small World by Dr Kirk Borne which is definitely worth your time. He also maintains a very active twitter presence and was voted the #1 influencer on Big Data by Onalytica.

11.) In a guest post on my blog at the beginning of November, Burtch Works’ entry-level recruiting specialists shared their advice in How to Get Your First Analytics Job.

12.) Researchers at the University of Pennsylvania analyzed social media Big Data to create word clouds showing commonalities by age, gender & personality. Fascinating stuff!

13.) Forbes published my article Five Ways Marketers Can Keep Quants From Quitting in December.

14.) We tallied results from our Flash Survey for analytics professionals about how many are approached via LinkedIn about new job opportunities and how often.

15.) What Design Thinking Can Teach Analytics Professionals was an attention-grabbing article from Data Informed.

With all the attention focused on Big Data, I expect next year to be even busier as more companies look beyond the buzzword and start seeing returns on their investments. The analytics hiring market will continue to heat up, so keep an eye out for my 2014 hiring predictions blog in the beginning of January. For more career advice, blog posts and industry news throughout the year be sure to follow Burtch Works on LinkedIn and Twitter. Happy Holidays everyone!

Kaggle Flourishes by Embracing the Competitive Nature of the Data Scientist

The emergence and success of Kaggle, a website dedicated to matching number crunchers with companies in need, shows us that no matter how challenging the problem there is always someone willing and eager to tackle it head on.

We’ve been hearing more and more about Kaggle in various capacities: candidates posting their Kaggle ranks on their LinkedIn profiles, students or recent grads using the site to gain experience with unstructured data, or even companies asking for candidates’ Kaggle rankings in job postings. For an aspiring data scientist with little to no experience this could be a invaluable tool to gain experience with real-world big, messy data.

Kaggle has grown in popularity from 25,000 to over 100,000 members in a little over a year and received attention from numerous publications such as The New York Times and Forbes about its unique service. Quantitative professionals from all over the globe can compete for prize money by offering intellectual property including any algorithms, models, solutions etc.

In addition to earning money for their services, Quants have the opportunity to hone their skills and see how their solutions and skills rank against other members. For companies who cannot afford or do not need a full time analytics staff, Kaggle will match you with a team based on your specific initiative that you can hire on a contract basis.  According to the founders of Kaggle, their intention was to inspire great performance by inciting competition amongst the members.

The competitions are divided into sections, with one dedicated to “Getting Started” which offers entries and students looking to test their skills on real data sets an opportunity to see how they compare to working data scientists worldwide. There have also been competitions from Belkin, Amazon and Expedia for more advanced participants.

Most recently, Facebook is using Kaggle to host a recruiting competition. Could this site completely revolutionize the way Quants learn, share, compete or job-hunt within the community? As anyone keeping abreast of Big Data news will tell you, we’re moving into the age of digital everything- socializing, interviewing, learning- why not digital sports for the data scientists too?

Data Scientists . . . or Data Wannabes?

Ladies and gentlemen of the analytic community – we have an epidemic. A seed has been planted in the minds of quantitative professionals that has sprouted and will not subside. The more we hear about Big Data, the more we hear about ‘Data Scientists.’ And like with any hot job, once the media starts to speculate on salary, candidates go nuts. As the President of Teradata Scott Gnau said in a recent article about careers in Big Data, “There are a lot of people who can spell Hadoop and put it on their resume and call themselves data scientists.”

Gnau makes a plea to get the term ‘Data Scientist’ defined and I could not agree more. When a national article makes the claim that one of these workers can be making $200K+ with a couple of years of experience, you can bet everybody who works with data will start calling themselves a Data Scientist. I have already seen it start to happen and thought I would offer my two cents on the reality of this highly coveted (and grossly overused) title.

The compensation issue is a tough one to tackle. Probably the most common question I get asked as a recruiter, by both candidates and clients, is related to salary. While it’s true that careers in marketing analytics are in demand and lucrative, this is all dependent on multiple factors including: years of experience; software familiarity; advanced degrees; pedigree of school; location preferred; and many more.  So when Rob Bearden, the CEO of Hortonworks, says that a “qualified data analyst” coming right out of school can make $125K, note that this is a very specific candidate with a very impressive background. To name just a few bullet points of the perfect candidate, data scientists typically have: 

• A PhD in Computer Science and advanced degrees in other highly quantitative disciplines like Mathematics  
• Knowledge of MapReduce tools like Hadoop
• Use of tools such as Python, R, Java, Hive, Pig
• Prior work or internship experience working with enormous, unstructured data sets

These are just general guidelines, of course. To be sure, a candidate with a Masters in Statistics can be a Data Scientist but this candidate will likely not be demanding as high of a salary as someone with two PhDs and Google on his resume. It might help to think of Data Scientists as a unique subset of the Big Data analytics profession. Just as an IT professional may work with Big Data, a marketing analyst who builds predictive models is not necessarily a Data Scientist.

Allow me to reiterate what I’m sure you have already heard countless times in the media: careers in Big Data are in demand and continue to grow. However, not every person with a quantitative background or experience in analytics is a real Data Scientist. The tools they are using are very new and many companies looking to take advantage of Big Data do not have a previously established foundation of Data Scientists to provide road maps for less experienced professionals. Likewise, innovative companies like Hortonworks that work to support tools such as Hive and HBase can and must offer workers competitive salaries to ensure the use of these systems are further developed.

For now we will continue to see ‘Data Scientist’ placed on resumes to catch the attention of hiring managers, and hiring managers will continue to shell out unprecedented resources to attract talent. Rightfully so – these guys are one in a million.