The metrics that make conversational banking work, with Forrester’s Aurélie L’Hostis

Banks no longer have a monopoly on financial guidance. Nearly as many consumers now ask ChatGPT about their finances as ask their bank’s own assistant, yet most banks still score their assistants on calls deflected. That metric can’t tell them whether customers got what they came for or whether those customers are quietly drifting elsewhere.

In this episode of Banking Reinvented, host Tim Rutten chats with guest speaker Aurélie L’Hostis, Principal Analyst at Forrester, who has tracked conversational banking since she wrote one of the first reports on chatbots in banking a decade ago. Drawing on that history, she explains how banks should measure conversational AI now that it’s moving from answering questions to helping customers actually complete tasks.

The conversation covers the four types of metrics she recommends, spanning customer outcomes, operations, business results, and risk; why governance belongs in the design from day one and works as an accelerator; how to organize ownership across a bank; and why, in 2027, banks will need to make their products visible and trustworthy to the AI agents customers use to find them.

Curious to find out more about AI in banking? Check out our content hub.