Primary tabs

For more than 50 years, banks have relied on computers and software to manage and secure their data, as well as protect their customers’ interests. We are now on the cusp of a major revolution—something that will be as big for banks as the internet was for retail businesses. What’s driving this momentous change? Artificial intelligence (AI) and advanced analytics, together with cloud computing, are enabling very large quantities of data to be collected, stored and processed in real time or near real time.

Before we discuss some of the opportunities to create new revenue streams and new business models through AI and analytics, let’s review a few general concepts that should be understood to ensure that, whatever you try to achieve with these new technologies, the results are likely to be successful.

Foundational steps in adopting AI and advanced analytics

Any AI or analytics project needs to start with a clear understanding of the desired outcome. What is it that you are trying to automate or otherwise achieve? You should tie this outcome to a clear vision of the dollar value. It should be clear, for example, how the project would deliver real return on investment for the business.

Consider an example use case where banks monitor and analyze business transaction data to establish when corporate customers are going to run out of liquidity. Based on large quantities of transactional and demographical data, the bank should be able to build competitive proposals for short-term loans and send them automatically to corporate customers. This process would increase new loan business, while delivering a first-class user experience for the corporate customer, driving satisfaction and loyalty.

Once you have a defined use case, along with clarity in terms of outcome, dollar value, ROI and success metrics, you can then select the right technique to deploy to achieve the desired outcome. Perhaps the right technique is machine learning. This technique allows you to gain expertise through experiences and complex patterns across very large data sets. It might involve, for example, building new algorithms to identify and analyze new fraud items in cross-border payment data.

Sourcing data for AI and advanced analytics

The next stage is to identify where you will source the data. Historically, organizations built big data repositories to drive reporting from a consistent view of the data. However, these repositories are too “after the event” to support machine learning and other advanced AI and analytics programs, which often need access to real-time data.

The other issue with big data is that it contains structured data, whereas AI and analytics can use unstructured data. With the latest tools, it may not even be necessary to move all of the data into a repository, but to access it at its source through the cloud. Here, organizations must ensure a comprehensive approach to cloud-based solutions that balances risk and value.

Many organizations are now building data lakes—repositories holding vast data—but for unstructured or raw data. These often are more suitable for machine learning and other advanced analytics programs. Once you identify the data and define the access, it is important to understand the necessary governance and regulatory issues around using the data. With more and more regulations defining how banks can use data, ensuring proper data usage is critical.

It can then be useful to build a proof of concept, a lightweight approach to validate the desired outcomes. This helps to build confidence in the approach and avoid costly mistakes.

Developing new revenue opportunities

AI and advanced analytics can provide significant benefits to banks, helping them to develop new revenue streams and improve the customer experience, as in the loan example outlined above.

Anti-financial crime is another area where banks have used machine learning for years to identify potential fraud cases or predict which “good” customers to add to a white list based on clear rule sets.  However, there now is the potential to enable new algorithms that predict new rules, enabling banks to stay one step ahead of ever-increasingly sophisticated financial criminals.

Other examples include the ability to understand customers better so that the bank can anticipate their actions at different stages of their lives, as well as better understand the customer experience and the variables that affect the customer relationship. This can provide valuable input for marketing tactics, including defining different segments of the customer base for targeted marketing, as well as input for the development of new products and service offerings.

Based on 2019 Global Treasurer News Transaction Banking Survey data, corporate treasurer satisfaction with their banking partners continues to decline significantly—a four-year trend. We not only see a four-year decline in satisfaction, but that decline continues to accelerate downwards. To succeed, leading banks are responding quickly.

What is a certain is that AI and advanced analytics are here to stay and will continue to automate and deliver yet unknown benefits to the industry. Feel free to contact me to discuss your own organization’s challenges and opportunities with respect to these advancing technologies. Also, learn more about CGI work with AI and advanced analytics.

À propos de l’auteur

Andy Schmidt

Andy Schmidt

Vice-président, Banques de détail

Andy Schmidt est un ancien banquier et analyste de l’industrie qui a comme mandat d’aider CGI à concrétiser sa stratégie au sein du secteur des services financiers. Il compte plus de 25 années d’expérience au sein des services financiers, d’abord en tant que banquier chez ...

Publier un commentaire

Comment editor

  • No HTML tags allowed.
  • Lines and paragraphs break automatically.
Règle de modération du blog et conditions d'utilisation