The huge demand for big data analytics, and scarcity in individuals with the right skills and experience, is driving the starting salaries for data scientists over $200,000, and it is going to get worse
A new species of techie is in demand these days—not only in Silicon Valley, but also in company headquarters around the world. Pascal Clement, the head of Amadeus Travel Intelligence in Madrid, says.
“Data scientists are the new superheroes.”
The description isn’t exactly hyperbolic: The qualifications for the job include the strength to tunnel through mountains of information and the vision to discern patterns where others see none. Clement’s outfit is part of Amadeus IT Holding, the world’s largest manager of flight bookings for airlines, which has more than 40 data scientists on its payroll, including some with a background in astrophysics. The company recently launched Schedule Recovery, a product that tracks delays and automatically rebooks all affected passengers.
A study by McKinsey projects that “by 2018, the U.S. alone may face a 50 percent to 60 percent gap between supply and requisite demand of deep analytic talent.” The shortage is already being felt across a broad spectrum of industries, including aerospace, insurance, pharmaceuticals, and finance. When the consulting firm Accenture surveyed its clients on their big-data strategies in April 2014, more than 90 percent said they planned to hire more employees with expertise in data science—most within a year. However, 41 percent of the more than 1,000 respondents cited a lack of talent as a chief obstacle. Narendra Mulani, senior managing director at Accenture Analytics, says.
“It will get worse before it gets better.”
Many data scientists have Ph.D.s or postdoctorates and a background in academic research, says Marco Bressan, president for data and analytics at BBVA, a Spanish bank that operates in 31 countries and has a team of more than 20 data scientists. He says.
“We have nanotechnologists, physicists, mathematicians, specialists in robotics. It’s people who can explore large volumes of data that aren’t structured.”
So-called unstructured data can include e-mails, videos, photos, social media, and other user-generated content. Data scientists write algorithms to extract insights from these troves of information. But “true data scientists are rare,” says Ricard Benjamins, head of business intelligence and big data at Telefónica, Europe’s second-largest phone company, which employs more than 200 of them. Stan Humphries, chief economist at Zillow, the real estate listings site, says:
“You can find a great developer and a great researcher who has a background in statistics, and maybe you can find a great problem solver, but to find that in the same person is hard.”
Universities are taking note. MIT, where graduate students in physics, astronomy, and biology are fielding offers from outside their chosen fields, is in the process of setting up a dedicated data-science institute. Marilyn Wilson, the university’s associate director for career development, says the center will begin enrolling graduate degree candidates in 2016.
In the U.K., the University of Warwick introduced a three-year undergraduate data-science program last year, which David Firth, the program’s mastermind, says may well be the first of its kind. He says.
“Big Business was complaining about the lack of people. Finance is a major employer, but also large-scale insurers, large online commercial retailers, high-tech startups, and government, which has huge data sets.”
Accenture’s Mulani says he’s tallied some 30 new data-science programs in North America, either up and running or in the works. The University of Virginia began offering a master’s in 2014, as did Stanford. Many of those students may be tempted to drop out before collecting their degree. Margot Gerritsen, director of Stanford’s Institute for Computational & Mathematical Engineering, says.
“Companies are scrambling. We have second- and third-year students getting offered salaries much higher than what I get.”
Starting pay for some full-time jobs is above $200,000, she reports. Summer internships, meanwhile, pay anywhere from $6,000 to $10,000 a month. To make these stints memorable, many employers offer perks such as free meals, complimentary gym memberships, and occasionally temporary housing. Gerritsen says.
“Sometimes you read about students getting abused in internships and working like slaves. We don’t see that.”
The bottom line: McKinsey projects that by 2018 demand for data scientists may be as much as 60 percent greater than the supply.
COMMENTARY: There is no doubt that data scientists are in high demand. There is simply not enough talent to fill the jobs. Why? Because the sexiest job of 21st century requires a mixture of broad, multi-disciplinary skills ranging from an intersection of mathematics, statistics, computer science, communication and business. Finding a data scientist is hard. Finding people who understand who a data scientist is, is equally hard.
Jean-Paul Isson, Monster Worldwide, Inc. says.
“Being a data scientist is not only about data crunching. It’s about understanding the business challenge, creating some valuable actionable insights to the data, and communicating their findings to the business.”
It is very unlikely that you will be able to hire a data scientist who can solve all your data problems. The skill-set presented below is a guide on how the modern data team should be equipped.
Click Image To Enlarge
What Makes a Great Data Scientist?
What are the personality traits of today’s data scientists? Do they have the skills needed to support businesses’ changing needs? How can senior management build cohesive teams to drive the most value from big data? Is this fast evolving discipline causing stress to its practitioners? These are some of the questions that SAS Analytics tried to answer when it conducted a survey in 2014 of 596 respondents in the UK and Ireland who identified themselves as part of the data science profession. The survey highlights 10 key personality types as well as some of the challenges data scientists are facing today.
Summary of Findings
- Data scientists with “traditional” traits (analytical, logical, technical) make up the largest group (41 per cent), yet a significant proportion of data scientists display other traits, such as strong communication and creativity skills.
- We expect a lot from data scientists: to be technically proficient, mathematically agile, business savvy and good at communicating. Yet 55 per cent of data scientists have fewer than three years of experience in the discipline, and more than a quarter are adapting their behaviours to fulfil roles that are not well matched to their skills or work personality profiles.
- In most personality types there is a 70:30 split between men and women. The Geeks, the profile with technical bias, strong logic and analytical skills, shows a higher percentage of women at 37 per cent compared to 63 per cent for men.
- Data scientists are exhibiting high levels of work-related stress, with a total of 55 per cent of the respondents showing a level of stress, with 1 in 4 male and just under a third of female data scientists being heavily stressed.
- Organisations must better identify and define what they need from data scientists. Only then can they build and develop multi-faceted teams with the complementary skills needed to realise the full value of big data.
This SAS survey report looked at the aggregated data of the initial 596 responses, received from 405 men (68 per cent) and 191 women (32 per cent). Respondents’ experience ranged from less than a year to a handful of data science pioneers who have committed more than 21 years to the discipline.
The survey was based on the well-established DISC profiling methodology, which has been used for more than half a century to categorise respondents into a range of personality types with recognised characteristics.
By using DISC, SAS able to create the typical profiles of today’s data scientists, and explore how these compare to skills and personalities required by organisations.
The DISC Personality Profiling Methodology
The DISC methodology has been widely used to assess personality types for decades. This information is often utilised to analyse and build effective teams, improve communication and increase productivity. DISC is based upon a two-axis matrix:
- Vertical Axis: Proactive to Reactive
- Horizontal Axis: Introvert to Extrovert
This generates four possible combinations – named D, I, S and C. Individuals possess measures of all four of these combinations but in varying amounts. The values for each combination is calculated and plotted on a graph to illustrate the relative strengths of each. These values can be matched to known behavioural characteristics to describe an individual’s personality profile.
Click Image To Enlarge
The Data Scientist Personality Mix
The SAS analysis identified ten psychometric profiles evident within the data scientist community, based on the responses to the survey so far. These profiles are characterised by definable patterns that are well known to scientists and were consistently apparent in the survey sample.
The ‘traditional’ traits associated with data scientists – such as technical, analytical and logical skills – still dominate. However, other less technical traits – such as project management, creativity and good communication skills – are also present. Organisations need data savvy individuals who are technically proficient, mathematically minded, business oriented and strong communicators.
It’s unlikely that any individual will have all of the skills required to maximise the value of big data. So it’s important for managers to identify the particular skills needed and build a cohesive team of individuals with complementary skills and traits. As we will see later in the report, failure to do so can result in individuals trying to fulfil roles to which they are not suited – which may lead to stress and burn-out.
Distribution of Personality Types in SAS Survey Sample
Click Image To Enlarge
Final Conclusions Drawn From The Survey
There is huge pressure on data scientists to display a range of skills: technical, mathematical, creative, business aptitude and communication. Unsurprisingly, organisations struggle to find this mix of skills in just one person, particularly when the majority of data scientists have only a few years of experience.
It is therefore important for managers to build cohesive, diverse teams that can answer all the needs of the organisation – not just data scientists but also data savvy managers to interpret the insights and transform them into actions. Organisations will need to:
- Manage the business’ expectations from data scientists
- Find and develop individuals with the appropriate skills and personality traits
- Create teams of individuals with complementary skills and experience
- Encourage peer to peer support, learning and development
- Match training to the business’ requirements.
It’s possible that, as technology develops, there will be less of a weighting towards the more traditional, technical personality traits – while demand for the less frequently found profiles such as the Ground Breakers, the Teachers and the Seekers will grow. As more respondents take part in SAS surveys, it is hoped to be able to explore this interesting development further.
Courtesy of an article dated June 4, 2015 appearing in Bloomberg Businessweek, an article dated August 30, 2014 appearing in Marketing Distillary and the results of a SAS survey titled, "What Makes A Great Data Scientist?"