Nine years ago, D.J. Patil and Jeff Hammerbacher coined the term “data scientist.” Four years later, Patil and Thomas Davenport deemed the profession “the sexiest job of the 21st century” in the pages of Harvard Business Review.
Ever since then, companies of all types have been struggling with a massive supply-and-demand imbalance of data scientists. For banks, this imbalance is particularly challenging at a time when the industry is increasingly vying for the same talent the tech giants are aggressively courting. To stay competitive, the financial services industry must defend against a new cast of competitors by changing how it attracts, hires and develops data science talent.
There are different types of data scientists but the purest definition of data scientist — the one referenced in the HBR article — is the research data scientist. The main job for this type of data scientist is to research new algorithmic techniques and write about them. With few exceptions, the research data scientist has a Ph.D. in statistics or something equivalent and possesses a range of skills — from math to coding. They are exceptionally hard to find and they are compensated accordingly. These are the kind the “Tech Five” — Google, Facebook, Apple, Microsoft and Amazon — employ.
The need for companies from all industries — including tech titans — to hire more research data scientists is only on the rise. However, the Tech Five have advantages in recruiting the elusive talent over banks.
For Ph.D. talent, the Tech Five are willing to pay handsomely — they will pay data scientists well beyond $500,000, and routinely, they will pay them $1 million per year. Truthfully, most banks are less inclined to make that commitment, which lets the Tech 5 stockpile talent.
But this dynamic must change in order for banks to mitigate the risk tech companies present. Indeed, the Tech Five have the scale, resources, ambition, and perhaps most importantly, datasets that challenge the world’s largest financial institutions. Unlike financial services organizations, the tech titan’s data isn’t necessarily siloed. Therefore, the Tech Five can put the data to use in ways that even the most sophisticated financial institution can only imagine.
While it is not a forgone conclusion that the Tech Five are trying to disrupt banking, it is a distinct possibility that these organizations will move past payments and into other components of the banking business.
If you privately ask bank CEOs what their top competitive concern is, they won’t mention another bank. They would mention Google or Facebook. It is the Tech Five, not the bitcoin startups, the robo-advisers or mobile-only banks that scare them.
Yes, there is a supply-and-demand issue for data scientists. But banks must battle harder for the talent that tech companies are impressing and shore up their own talent in these two ways:
Commit to compete
Finance is a rich area for research data scientists. Banks need to find the talent that wants to work on their problems and actively engage with them. Already, banks run fintech competitions. But they should also run data science competitions.
These competitions should target students in the more than 100 data science programs that exist at the university level. Once banks have identified the talent most interested in the subject matter, banks need do whatever it takes to get them onboard. This means paying at the same level that Google and Amazon do, pairing them with mentors in the bank and giving them time to do original research.
Invest at every level
The inverse of aggressively hiring is developing the talent that is already in the bank. Banks can and should nurture data science talent in house that will understand the business, the culture and the math. Put these individuals at the center of the data science effort and give them the exposure and authority within the institution that will keep them there. It will not be long before these data scientists are on the board of the world’s largest banks.
The claim that the data scientist is the “sexiest job of the 21st century” is audacious. A century is a long time and the acceleration of change will challenge that assertion. What remains clear, however, is the fact that the next decade will be pivotal for global systemically important banks to make sense of their data as AI irrevocably changes the competitive balance.