The Spanish bank reported encouraging results with its pilot of an AI-based solution from research lab OpenAI that analyzes customer conversations about banking. BBVA tested the model on 10 conversations taken from online forums, with the goal of generating summaries for staff to read—saving them the time and effort of evaluating individual conversations—and act upon. The solution identified each conversation’s core issues and questions, what product or service was involved, and the recommendation or answer given and its status of completion. While accuracy varied across the small sample size, BBVA said the results show great promise, and that it will further explore the model’s value.

By leveraging AI, banks can derive insights from customer behaviors and apply them to personalized product offers and recommendations. AI solutions can process and evaluate vast stores of data more efficiently than human employees, freeing up valuable resources to instead go toward applying insights to product strategy. BBVA has clearly recognized this value, as the OpenAI trial is the latest in a string of AI-related moves at the bank: It launched its global AI development center, AI Factory, in June 2019 and hired hundreds of new data scientists and AI experts to better understand its clients’ needs and offer them tailored products. The bank has since seen rising digital sales and conversions through AI-driven marketing campaigns. While the customer conversation evaluation tool still needs fine tuning for accuracy, it has the potential to improve BBVA’s customer service and bolster its ability to offer relevant information to customers—which could ultimately lead to more engagement and greater satisfaction.