During the 1960s, it was the hand-held calculator. The ’90s brought the worldwide web. The hottest innovation of the 2010s? The mysterious field of data science is surely nearing the top of the list.
Five years have now passed since Harvard Business Review famously declared data scientist “the sexiest job of the 21st century”. The accuracy of this label has been much debated but what is far less contentious from today’s perspective is how critical a cog data analysis has become in the advertising and marketing machine, as in so many others.
As Jon Mackay, MD of MRM McCann Hong Kong, puts it: “If an organisation in this industry does not have a strong data, analytics and performance strategy interwoven into the fabric of the broader organisational global strategy then they will be at a distinct disadvantage against their competitors that have.”
The research group Forrester predicted in a recent survey that the market for big data management solutions will grow by 12.8 percent a year for the next five years. The US is still the world’s largest market — but Asia-Pacific is the fastest-growing.
Salaries for the job du jour are skyrocketing as a result, even at junior level. India’s International School of Engineering (INSOFE), which offers a popular data analysis certificate in Hyderabad and Bangalore, reveals that its average starting salary for graduates is 650,000 rupees (USD$9,740) — but that students are reporting a 300-percent hike by their second year in work. Junior data scientists in Singapore, meanwhile, earn an average annual salary of S$59,480 (US$42,240), according to a report by recruitment firm Big Cloud; this is already higher than the country-wide mean wage.
All the major players in Asia’s media and marketing community are placing big bets on data science, enthralled by its ability to shine a light on customer behaviour with more precision than ever before.
There’s a problem, however. The world has a massive shortage of data scientists, and this holds as true in Asia as in Europe or America.
The recruitment specialist TeamLease predicts a demand-supply gap of 200,000 analytics professionals over the next three years in India alone — and India is widely held to be further ahead than most countries in terms of data science educational opportunities. Other markets in the region, such as Hong Kong, are much less mature, experts say.
“Right now, it’s about motivating people to get into the field and unfortunately Hong Kong is quite lagging behind,” says Ivan Lam, associate director of technology and operations at Morgan McKinley. “Do we have the chance to be a really hot competitor in the region on the data science remit? We still have a long way to go.” The data scientists who do find roles in Hong Kong are often recruited from overseas, he says.
(Story continues below)
DATA STUDENTS DISCUSS THEIR SUBJECT
L-R: Top to bottom below
Ryan Ng, 22
Vinita Phadke, 25
Dibyaroop Samajpati, 31
Amogh Mannekote, 20
Hitesh Mariyala, 27
Ooi’s counterparts are facing similar difficulties in recruitment. “Data science is actually a very broad set of skills which spans statistics, database management, automation and machine learning, to name a few,” says Naman Sharma, chief analytics and insights officer at Havas Media Group. “It is practically impossible to find a data scientist with all the required skills.”
Mackay agrees, saying that the perfect data analyst for the marketing world is a rare and complex beast. “A good data scientist not only has to be able to manage large sets of data from different channels or data capture points, but they need to understand the significance of each data set and the role and weight it will play in its eventual application.” Beyond this, he continues, they need to be able to optimise the performance of these models to build a strong understanding of the customer and their behaviour, in order to help brands build more meaningful relationships with them. Finally, they need to be client-facing, personable and have a unique ability to make “the complex simple and the simple compelling”.
“Finding all three of these skillsets can be challenging,” he concedes.
The problem these companies face is likely to get worse before it gets better, given that educational institutions across Asia are only just beginning to respond to demand.
India is probably the fastest-moving country in this respect. While INSOFE was particularly forward-thinking in offering its six-month certificate programme in big data analytics and optimisation as early as 2011 (the first study cohort contained just 12 students; now there are 350 studying at any one time and prospective joiners face a months-long waiting list), plenty of other courses and degrees are establishing themselves across the country’s tech-forward cities.
“India is flooded with talent”, according to Morgan McKinley’s Lam. “But China will also be a key competitor in this space. Not just because they have their own talent pool already but also because they are also attracting people from all different places to come to them.” In the last two years, new courses have sprung up in Beijing’s Renmin University and Zhejiang University in Hangzhou; and INSOFE’s management team have recently been approached by the Chinese to start developing programmes there too.
Singapore, meanwhile, is becoming another exciting hub for data science learning. The first data science degree in the region launched in August, a four-year course at the National University of Singapore, and Data Science Dojo — a US-based programme that offers a bootcamp-style introduction to data analysis — is also running two courses a year in Singapore with plans to set up in Bangalore in the future.
The students currently taking these courses — whether they are fresh out of school or transferring to data from a different career — seem likely to be the stars of industry tomorrow, putting them in a position of some power. So what do they plan to do with their high-worth degrees? All those interviewed by Campaign Asia-Pacific see data science as a job for life. Being in their 20s and early 30s and therefore digital natives, they describe their field of study as a “passion”, something they are genuinely excited about.
Their career desires appear to fall into two main camps: students who want to work in finance, where starter job offers abound and salaries are highest; and those who want to use their skills to make a difference in society.
Marketing or advertising isn’t high on their agenda yet — and Professor Vinesh Thiruchelvam, dean of the Faculty of Computing, Engineering and Technology at Malaysia’s Asia-Pacific University, agrees with Ooi that one explanation could be a lack of sufficient noise from the media industry about how diverse and evolved it has become.
“Marketing companies seem to be the quieter group in this data science frontier,” he comments.
This situation should be changing as companies such as Havas and Dentsu Aegis continue to pour resources into pioneering data centres in the region, equipping them with the resources to train up a new generation of data bods.
Even so, if they don’t get the right grades all the open doors facing these students will slam shut, say their professors. “Those who do very well are being lapped up,” says Dr Sridhar Pappu, executive VP of academics at INSOFE. “They are getting multiple offers. At the same time, the field is such that even if there are a lot of positions available, if you don’t perform well, people are not filling them just because there are vacancies.”
This perhaps goes to the heart of one of the biggest misconceptions about data science: that it can easily be applied to effect by anyone with a little training. “People can come to very different conclusions from the same data,” warns Pappu. “If someone just takes training in one tool, they will input some data and run some techniques and an output will come, but that output may be completely junk. Unless you know the context very well, you cannot do that.”
The field is also changing so fast that it requires serious dedication to keep up. “In all probability the skill you most covet today will be completely irrelevant in a couple of years,” Havas’ Sharma points out. “Hence, the most important ability in a data science practitioner is to self-learn and to always be on the lookout for new technologies.”
So with robots and artificial intelligence predicted to steal jobs across the business arena in coming years, is data science a future-proof choice for ambitious and dedicated youngsters?
A recent Gartner report estimated that 40 percent of all data science tasks will be automated by 2020. But here again, data scientists may have the last laugh: yes, machines and computer programmes may get more intelligent — but they also need to be taught how to get more intelligent. Those with the power to instruct them look likely to hold the reins of the future.