Conversational artificial intelligence (AI) has been one of the hottest areas in fintech for nearly a decade, with an array of startups executing on a vision originating from 1950s science fiction. Initially, the “conversation” was limited to text chatbots, but more entities, including financial institutions, have implemented voice-enabled chat, offering real-time responses and support for queries made over the phone and through devices like Alexa and Google Home.
Interest has increased in recent months, as the pandemic has made reducing costs and improving convenience very important to service providers. Additionally, staffing challenges and an increase in remote banking have led many FIs to revisit investments in AI after spending time on the sidelines.
For those FIs expressing interest, it’s good to review some of the lesser-known factors underpinning a successful implementation of conversational AI.
Powering through the Great Resignation
A persistent barrier to any AI project is employee suspicion that the main objective is eliminating jobs. Despite ample evidence that the technology frees staff from menial tasks, letting them focus on higher-value ones, concern about potential layoffs often leads to subtle roadblocks. These objections can be overcome, but they require collaborative leadership and solid, transparent communication.
Add in the Great Resignation…frontline workers, especially contact center reps and tellers, feel the pain of short staffing. Employees are more likely to embrace the relief AI can provide as it becomes clear the tech will not lead to massive layoffs.
The most significant pain points are in contact centers as traffic shifts from branches to remote channels. There’s been an increased emphasis on voice AI to offload routine queries, which is more effective than legacy phone interactive voice response (IVR) systems.
Speaking of IVR, some institutions are cutting costs by replacing existing systems. However, it must be stated that while voice and text programs are valuable, FIs cannot expect them to cure all ills immediately.
Learning by Doing
While several solid conversational AI vendors exist, they are not interchangeable. Each has specific strengths; FIs need to select solutions best aligned with their unique situations, immediate needs, and long-term aspirations.
Banks and credit unions must understand that conversational AI is not a “plug and play” technology that automatically delivers results once installed. Institutions can pick the exact solution for the same reasons yet have very different outcomes. This is due to how implementation is handled, including employee training protocols and the software itself. Easily overlooked nuances such as demographic and geographic differences can create obstacles with voice queries tied to regional terminology, accents, and dialects.
Vendor collaboration to articulate and design the desired client experience is essential to success. Robust customer journey maps and segmentation that allow for highly tailored chat experiences will be the difference between revolutionized service and situations where clients dismiss a solution as “another dumb chatbot” and immediately try to speak to a human.
Excelling at conversational AI often requires dedicated internal resources to manage, monitor, and enhance the offering – tweaking the training data and responses, expanding transactional capabilities, and rapidly adding support for new request types. With the right tools and focus, conversational AI can be a fantastic channel for client communication and data mining, lending itself to A/B testing for promotions and collecting feedback for products and services.
The Bottom Line
Conversational AI was an important technology solution to augment FI toolkits prior to the pandemic. The challenges posed by the Great Resignation and an accelerated shift to digital channels have brought its value into even greater focus.
As customer demographics continue to shift, quality 24/7 self-service options will increasingly be table stakes for banks and credit unions. Digital modernization is an existential necessity – a matter of when, not if – and many FIs have already started that journey.
SRM has found that success with AI depends not only on selecting the vendor that best matches an FI’s needs but also on executing a well-considered plan for implementation and continuous improvement. SRM can help you evaluate your plans and an appropriate course of action.