In this installment of “Questions From Our Last Meeting,” we cover a common question that we receive from financial institutions that are considering enhancing or updating their digital infrastructure surrounding what technology, if any, needs to be in place to realize the best value from a Virtual Financial Assistant.
Setting the Scene:
Date: October 2020
Financial Institution (FI) Type: Credit Union
Size: $1 billion-$5 billion in assets; > 100,000 Members
“Do we need live chat or other technology installed before offering the Abe.ai Virtual Financial Assistant (VFA) to our members?”
- It’s based on the current situation and preference of the FI. In general, an FI doesn’t have to have any specific technology, including live chat, to implement our VFA. We can integrate it into the FI’s existing mobile app or web banking experience.
- If your FI already offers or is planning to implement live agent chat as part of your customer service capabilities, then it does make sense for this technology to be in place first. While our AI layer “sits in front” of the live-agent chat and messaging capabilities, we plug into existing providers of those services to provide the option for a seamless transition from the VFA to live agent. We facilitate this seamless handoff to a human at the first indication that a customer would prefer human-to-human interaction instead of working with a VFA.
- If your FI doesn’t have or plan to offer live chat, but still wants to be able to transition from the VFA to live customer support, we can also script to a dynamic phone number that lets members/customers simply click to start a phone conversation. The nice part in this scenario is that you don’t have to train agents to use a new chat offering, so there’s no need to re-forecast resources. Your FI also doesn’t have to invest in more technology before adopting AI, so this is the most cost-efficient approach without sacrificing service.
- Regardless of the preferred architecture, data from our clients show that more than half of all requests are successfully resolved via the AI layer without the need for human involvement in the first month – and that number increases to close to ninety percent by month six.