The future of data-driven banking isn’t data scientists replacing bankers. It’s bankers harnessing technology to generate the insights they need for themselves
Data culture has become a huge focus for data leaders in recent years. It’s clear that data leaders will only be able to maximize the value of their companies’ data assets if they can get frontline staff to adopt new, data-driven ways of thinking and working. But the ‘end game’ of all these cultural change initiatives can differ from company to company.
For Manav Misra, CDAO at Regions Financial Corporation, the ultimate goal is to convince the whole organization about the value of data and analytics.
“When I came into Regions, my vision statement for my organization was to make Regions a best in class data-driven organization,” he explains. “To do that, I had to not only build my team, but also to get the rest of the organization to believe that data could really address their problems.”
However, other data leaders have grander ambitions. For Dan Costanza, Chief Data Scientist at Citi’s investment bank, getting the whole business on board with the data project is just the beginning.
“Before you see investment bankers’ jobs getting replaced by data scientists, I think the data scientist jobs are going to be replaced by investment bankers,” he quips. “If you think about what makes someone good at my job, it’s really the same thing that makes someone good as an investment banker.”
“What you’ll see is that the banks that are effective at using data science will be the ones where the bankers themselves are the data scientists”– Dan Costanza, Chief Data Scientist, Citi
Costanza notes that much of his job involves taking existing solutions to problems and applying them to new and interesting things. For him, the hardest part is understanding the business problems at hand, knowing the different ways you can approach them and figuring out how to apply the tools in his toolkit to them effectively.
“More and more you’re going to see a lot of the data science toolkit will have to become one tool among many in a banker’s toolkit,” he predicts.
Self-Service Tools Will Democratize Data
One huge challenge when scaling advanced data or analytics capabilities is that it takes a great deal of knowledge to write and work with code.
Of course, it also takes domain expertise to apply data-driven techniques in a business context – and that’s why it’s so important for data teams to work closely with the business units they’re serving. But Costanza argues that this division of labor may not be necessary in the future, as data literacy improves.
“If you ask the question of a first-year analyst coming into a bank,” Costanza notes. “More than half of the people coming in know how to code. There’s no reason you couldn’t have them doing something that’s one step shy of coding or interacting with existing code and playing around with it.”
The promise of self-service data technologies is that they will arm ordinary staff with the tools they need to ask and answer their own questions.
“Technology is never going to replace humans,” explains Mike Kim, CTO and Co-Founder of self-service AI platform Outlier. “But how do we help them do the human thing that they do better?”
Kim argues that a true self-service data solution should empower staff members to investigate multiple data sources, detect data quality issues, generate data visualizations, build and productionize data models and communicate useful insights to the relevant stakeholders.
“Whatever self-service solution you find has got to be able to fit that shape,” he says. “It has to be able to turn over every possible stone that’s left there and figure out what’s important in a business context.”
“There’s no reason you couldn’t have them doing something that’s one step shy of coding or interacting with existing code and playing around with it”– Dan Costanza, Chief Data Scientist, Citi
As self-service data technologies become increasingly sophisticated, frontline staff will become able to handle many of the tasks they depend on data scientists for today. At this point, it will become possible to scale data-driven capabilities across an organization without needing to hire more expensive data scientists. The valuable insights that can be unlocked with data will be in the hands of the many, not the few.
The Road to Data Democratization
Data democratization will remain something of a pipe dream for CDAOs at companies where demand for data-driven capabilities is weak.
As many data leaders have learned the hard way, simply telling people what can be done with data isn’t enough. As Misra says, frontline staff need to see the benefits these new ways of working can bring for themselves before they will truly buy into them.
“Deliver success early,” he recommends. “That’s created such a buzz overall that now I don’t need to go and sell anymore. There are people coming to me with lots of requests of what they want done.”
Costanza agrees that the best approach is to develop new capabilities for the business units that are most passionate about data first.
“My advice to people on that is always to start micro,” he says. “If you try and go and convince the whole organization that they should be doing these things, that’s very hard. If you want to go and convince a couple of people of it, that’s a whole lot easier.
“And if you’re able to do that well enough that they start to talk about the things you’re doing and bringing to the table to others, that’s way more effective than trying to start from the top down.”
Edgar Abreu, VP, Data Analytics at financial services company Synchrony also knew he would need to deliver successes quickly when he first joined the company. He says appointing ‘data champions’ within the teams he liaised with played a key role in changing the way they think about data.
“Those data and analytics champions would advise us, work with us and tell us when those audits were starting,” he explains. “Basically, by winning them over, they helped us win the rest over.”
With the right ecosystem in place for data culture to flourish and grow, demand for data-driven capabilities will begin to spread throughout the organization. Then, the advent of self-service data platforms will empower frontline staff to truly embrace data-driven ways of working.