BBVA Contact Centre AI Agents Cut Handling Time by 15%

What happened

BBVA's contact centre teams in Italy and Germany — the banking group's two fully digital retail banks — have independently developed and deployed a pair of generative AI assistants designed to help agents handle customer enquiries more quickly. Unusually, the tools were built by the operational staff managing those contact centres rather than by a centralised technology function, signalling a shift in how the bank approaches front-line innovation.

The AI assistants surface relevant information about products, services and standard procedures in real time, allowing agents to respond to the most common customer queries without needing to search multiple systems or escalate. The result, according to reporting by Finextra, is a reduction in average handling time for those enquiry types of more than 15%.

Why it matters

Average handling time is one of the most closely watched metrics in contact centre management, but it has always sat in tension with quality: pressure to resolve calls quickly can erode the depth of care an agent provides. What makes this deployment notable from a CX and behavioural standpoint is that the efficiency gain appears to come from reducing cognitive load on the agent — freeing attention that can, in principle, be redirected toward the human dimension of the interaction rather than information retrieval. When agents spend less time hunting for answers, they are better positioned to listen, empathise and problem-solve.

The fact that the tools were conceived and built by the people who actually run the contact centres, rather than handed down from a product or engineering team, is also significant for service design. Frontline staff carry tacit knowledge about failure points and friction that rarely surfaces in top-down technology projects. This model of bottom-up AI development points to a broader design principle: the people closest to the customer experience are often best placed to identify where automation genuinely helps rather than merely displaces.

By the numbers

• More than 15% reduction in average handling time for the most common customer enquiries, achieved across BBVA's Italian and German digital bank contact centres.
• Two generative AI assistants developed and deployed, one per market, by the local contact centre management teams.

The Renascence take

Most coverage of AI in contact centres focuses on chatbots replacing agents or on cost-cutting headlines. BBVA's story is subtler and, for that reason, more instructive: it is about augmenting agent capability from the inside out, with the people who feel the friction designing the fix.

The behavioural principle here is cognitive offloading — reducing the mental effort required to retrieve information so that agents can invest that reclaimed capacity in higher-order tasks like de-escalation, empathy and judgement. A 15%-plus reduction in handling time is meaningful, but the more durable gain may be in agent confidence and consistency. Customer-obsessed operators should take note of the governance model as much as the technology: when frontline teams own the tools they use, adoption is higher, iteration is faster, and the resulting solutions are far more likely to address real friction rather than assumed friction. The question worth asking internally is not "how do we deploy AI in our contact centre?" but "who in our contact centre understands the problem well enough to design the solution?"