Superior origination performance: The key foundational strategies for digital banks
For superior origination performance, combining artificial intelligence (AI)/machine learning (ML) models and automation are key foundational strategies for digital banks. AI/ML, and automation are often treated as separate value chain transformation initiatives. A holistic end-to-end approach to benchmark and transform the value chain with a combined approach can radically deliver both superior origination performance and improved customer experience. Insight feedback loops from intelligent banking models can be used to rapidly enhance digital engagement, banking product personalization and automate operating processes.
The spike in demand caused by the recent pandemic necessitated even greater attention on initiatives to increase business model performance. Banks have accelerated and expanded digital investments in both origination process automation and embedding the use of AI/ML. These investments have been made to increase customer acquisition rates, enable differentiated pricing, optimize profitable credit risk management and automation to eliminate on-boarding friction. The demand for AI/ML based Intelligent banking is growing with an expanding range of use cases. The market is expected to grow to c$64bn by 2030 from c$4bn currently, a 32% CAGR over the period.
The demand for AI/ML based intelligent banking is growing with an expanding range of use cases. The market is expected to grow to c$64bn by 2030 from c$4bn currently, a 32% CAGR over the period.
Two interesting origination value chain user cases illustrates the potential business value:
1) Targeting personalized products and services together with the right engagement channels using an expanded universe of data sources generates business value.
In a recent example with a European regional bank, Atos AI-driven tool Persona 360° was used to drive higher customer engagement by building a customer lifestyle and digital engagement profile to identify more effective engagement strategies and likely product affinities.
A significant percentage of the customer base was identified as ‘limited relationship and engagement’ with less than 4 contacts per year and an affinity for investment products. The new strategy was used to implement a personalized digital marketing campaign for the target segment, leading to a 35% increase in conversions and 37% higher engagement.
Product on-boarding that removed key friction points by using automated KYC / AML (Know Your Customer / Anti-Money Laundering) checks against government and other databases, together with verification of funding sources, enabled customers to be onboarded in a matter of minutes.
2) The growth objectives of banks can be enhanced by ensuring AI/ML model design incorporates fairness outcomes whilst using digital data sources to enrich scoring algorithms.
Fairness in access to financial products is a key aspect of banking regulation and conduct. Embedded bias excludes otherwise suitable and profitable customer segments. Credit risk processes can utilize AI/ML challenger models to test for bias both in the datasets and the algorithms themselves.
Additionally, products designed for traditionally underrepresented segments can further lift origination rate performance. For example, the effectiveness of this strategy is illustrated by WeBanks’ successful growth strategy of serving the unbanked.
Atos combines AI/ML expertise from specialist AI/ML business such as Atos DataSentics with Banking Platforms automation to enable banks to effectively combine these technologies.