AI can translate into meaningful outcomes with a focus on personalisation

While AI is still far from replacing human contextualisation or emotional intelligence fully, the use of AI to even aid opinion is still significantly underutilized

The cool factor and buzzword status of Artificial Intelligence (AI) aside, there is a genuine case to be made for leveraging AI in solving one of the biggest focus areas in the Indian retail banking today - making the transition from product focus to customer centricity.
 
However, the reality remains that except for a few use cases such as chatbots and back-office automation, AI’s foray into Indian retail banking is yet to truly succeed in translating that magic sauce into a personalised experience for customers or material business result for the banks.
 
I will attempt to explain four key enablers that banks can provide to ensure that their AI efforts translate into meaningful and material outcomes with a focus on personalisation and enhancing client experience.
 
Structural enablement : The heart of the structural issue is that often AI is just another initiative within the digital initiatives or data science departments rather than the core of business decisions. This has led to a lot of AI projects becoming use case and department-specific rather than a conscious attempt to rethink and re-engineer the entire end-to-end client journey. Enabling this structurally is not easy and much effort needs to be invested in changing how we manage of not just the AI processes but the entire organisation’s thought process. To achieve this, we need a potent mix of continuous education, evangelisation, bottom-up feedback, up-skilling and re-skilling across all functions and departments.

Data enablement: In AI, the algorithms are just the cooking process but finally it's all about the quality of the main ingredient -data. Solving for data architectures and availability is the single most important pre-requisite of any large AI implementation programme. From knowing how much you spent in that bar last Saturday night to how much your electricity bill was last month, the banks know everything. This data is from actual transactions and hence factual. Yet clean, formatted and usable data is still a challenge in almost all banks. Interestingly AI itself can play a critical role in cleaning up and enriching the raw data to make it ready for deeper AI insights in a virtuous cycle.

Raghu Mohapatra, Group chief operating officer, Goals 101 Data Solutions

Process enablement: Would you ever buy a shirt just because it was blue?  Clearly not! The price, the brand, the tailoring design and the fit all need to appeal to you in tandem to make that final purchase decision.
 
So why should it be any different with respect to banking cross-sell? Banks in India have traditionally used data science to build propensity and risk models focused on answering "what" to offer to an existing customer, that is, the proverbial 'next best offer'. But this is just one axis of the full personalisation experience.
 
As individuals, we also have distinct personal preferences with respect to the channel via which we want to consume banking communication (e.g. SMS, e-mail, digital etc), as well as content that appeals to us and the time in which we would prefer to consume it. The "when (time of sending)", "how (dynamic personalised content)" and "which (channel)" are often left out of the AI-led personalisation, reducing the true potential impact of the communication.
 
Cultural enablement : The prioritisation of traditional gut instinct and domain experience over AI is still a challenge when it comes to applying AI to front-end client engagement use cases. While AI is still far from replacing human contextualisation or emotional intelligence fully, the use of AI to even aid opinion is still significantly underutilised.