Focus On AI

Banking On AI: Societe Generale’s Journey

A half a billion euros. That’s the amount of run-rate value creation that Societe Generale aims to generate from data and AI by 2025, Julien Molez, the Group Innovation Data & AI Leader, told a gathering of executives at the AI For Finance 2022 conference in Paris September 20.

The potential for value creation for banks is one of the largest across industries, as AI can potentially unlock $1 trillion of incremental value  annually, according to a McKinsey report.  

Across more than 25 use cases, AI technologies can help boost bank revenues through increased personalization of services to customers and employees; lower costs through efficiencies generated by higher automation, reduce errors rates, and improve resource utilization; and uncover new and previously unrealized opportunities based on an improved ability to process and generate insights from vast troves of data, according to the global management consulting firm.  More broadly, McKinsey says disruptive AI technologies can dramatically improve banks’ ability to achieve four key outcomes: higher profits, at-scale personalization, distinctive omnichannel experiences, and rapid innovation cycles. Banks that fail to make AI central to their core strategy and operations—risk being overtaken by competition and deserted by their customers.

The value is clear and the pressure to succeed is strong but getting there is difficult. Banks and large corporates in other sectors are all struggling to diffuse and scale artificial intelligence and reap the full benefits of the technology. Societe Generale’s journey gives some key insights into how traditional companies can begin to build effective AI strategies.

Developing A Lingua Franca

Until about four years ago Societe Generale’s, focus on data was mainly geared towards managing increasing regulatory pressures (i.e. BCBS239, GDPR, etc.) and handling data quality and documentation issues. Active usage of data was still mostly a technical topic handled by data experts within IT department with limited business acumen. “Although we had a rich data patrimony, we were not a data-driven company,” Molez said in an interview with The Innovator. With the CEO’s support, an internal team, working closely with the Group Chief Data Officer, was set up to change the way success was measured and focus on use cases that generate business value. “It is not an easy switch,” he says, “It requires a lot of change management but unless you align with the business, have a sponsor and a clear vision of the outcome, it is not worth spending the time and money,” he says.

 Societe Generale created a value framework and purposefully oversimplified it to guarantee that people working on the business side of the bank’s operations would feel comfortable using it. The form had ten fields, with a dropdown menu. It asked business users to propose an AI or data use case and identify what business problem it was meant to solve and the pain points. They were also asked to quantify the expected value of a successful outcome. There were three choices: over €1 million, between €200,000 and €1 million or under €200,000.

“The focus on value enabled us to develop a global portfolio use case at the group level,” says Molez. “It was a big step that engaged all the stakeholders and allowed us to generate one global value figure for the group, using one framework, one platform, one governance system and one vision of all initiatives, with a dashboard showing the progress plus group indicators and targets,” he says. “It is very powerful to speak the same language.”

A Data Strategy Based On Four Key Pillars

Based on Societe Generale’s experience Molez recommends building a data and AI strategy based on four key pillars. “You must invest in all four simultaneously,” he says. The first is strategic alignment with the business, the second is data management and data quality, the third is a state-of-the-art tech platform, the fourth is skills.

“People are key to scale,” says Molez. The bank made some errors early on, he says. “Our mistake was thinking that data scientists were the most important part of the value chain,” he says.  “It is a critical factor, but not the only one.” If a company’s AI strategy is not ready and the right stakeholders are not lined up, data scientists are going to end up working alone, will not be able to progress, and will wind up leaving, he says.  To attract talent first a company needs to create an appropriate workflow for data scientists, says Molez. It may not be necessary to recruit the people at the top of the field. “It is like football and baseball, there are different leagues,” says Molez. “Companies need to be honest with themselves about what league they are in.”

Who you hire is not the only criteria. Where they are located is also key.  Societe Generale embeds its data scientists in business units.  And, to optimize interactions with data scientists, the bank is training its top business managers to generate value out of data. Some 600 have received some training and more than 120 have been signed up for more in-depth training in courses at Harvard Business School and HEC.

The bank is intent on getting the most of its data but believes that machine learning is only one of multiple paths. Basic analytics enable the company to create dynamic real time dashboards and have managers in the organization look at them before making decisions, he says. Big Data is important because it underpins high performance computing. Technologies such as PET (privacy enhancement technologies) that allow the sharing of data without revealing competitive information or violating privacy rules, is another important area to develop, says Molez.

The Next Frontier

Today Societe Generale is applying AI on numerous use cases including classic ones like payment fraud detection, chatbots or more innovative ones on the field of capital market predictions. The potential is much greater, says Molez.

The next frontier is personalized services that are relevant and timely, and based on a detailed understanding of customers’ past behavior and context. The bank is proceeding cautiously because on the one hand customers say they want more personalized services, but they also inherently regard banks as trusted third-party that safeguard very sensitive data.

It has taken a concerted three year effort to get the bank where it is today on its AI and data development path. “It is an infinite journey,” says Molez.  “It requires deep changes to the organization if you want to use AI at scale,” he says.  “It takes time. You need to change the way of doing business and change skill sets. It is going to take more years and not just one or two to make those changes.”

This article is content that would normally only be available to subscribers. Sign up for a four-week free trial to see what you have been missing.

About the author

Jennifer L. Schenker

Jennifer L. Schenker, an award-winning journalist, has been covering the global tech industry from Europe since 1985, working full-time, at various points in her career for the Wall Street Journal Europe, Time Magazine, International Herald Tribune, Red Herring and BusinessWeek. She is currently the editor-in-chief of The Innovator, an English-language global publication about the digital transformation of business. Jennifer was voted one of the 50 most inspiring women in technology in Europe in 2015 and 2016 and was named by Forbes Magazine in 2018 as one of the 30 women leaders disrupting tech in France. She has been a World Economic Forum Tech Pioneers judge for 20 years. She lives in Paris and has dual U.S. and French citizenship.