Later this year, some banking customers may be able to get their questions answered by IBM's Watson. As part of a new strategic partnership, the Genesys Customer Experience Platform will be integrated with the IBM Watson Engagement Advisor. The integration will benefit agents and customers alike. A self-service customer will be able to get his or her answer via Watson, and call center agents will be able to use Watson, instead of rule-based decision trees, to search for answers to customers' queries.
"This partnership marries the cognitive capabilities of Watson with a well-entrenched domain expert in the contact center space," describes Stephen Gold, vice president of IBM Watson Group. Genesys, which provides contact center and customer experience software, has 4,500 customers in 80 countries. What's more, Gold expects that some of those customers will experience Watson this year, less than four years after the computer was thrust into the public eye when it beat a human on "Jeopardy!."
Genesys, which has a strong foothold in the financial services industry, will be making Watson available to banking customers in the fourth quarter of this year. A demo of the product showcased at Genesys' G-Force conference was "conceived, built, and integrated in three weeks," Gold reports, suggesting that Watson will not be time-consuming to add on to existing contact center operations.
Among Watson's theoretical benefits is the ability to reduce agent churn. "For an agent, I think this is a game-changer," Gold says. "Turnover for agents in this business is relatively high, and the big reason is their frustration in their inability to do their job well. Watson puts at their disposal a more intuitive, natural interaction."
Today, agents often look up responses by going through decision trees or rules-based approaches. "That only works well if the line of discussion with the customer is aligned with that logic," Gold says. Watson can offer a few different answers to a question. An agent can choose the answer based on Watson's stated confidence in each answer. If a response isn't helpful, the agent makes a note, and machine learning takes over to improve the results over time.
For the banking industry, Watson could answer questions ranging from "How much will it cost to send my daughter to college?" to "I missed my mortgage payment. Can you waive any late fees?" In the latter case, Watson might alert a client to a policy that waives a first-time fee, for example.
Companies using Watson could also set up rules that would help retain clients. "There could be areas where Watson could answer, but shouldn't," Gold says. If someone asks Watson, "How do I cancel my credit card?" he could be trained to ask why, or a person could be transferred to a live agent who would try to prevent the customer from closing their account.
These experiences could take place through a Web chat session, smartphone, or tablet. There is also potential for Watson to work on a phone, if a company uses a speech-to-text technology that turns the verbal questions into words for Watson to read.
The Watson integration will also ensure that consumers receive a consistent experience across channels. A company using the Genesys-Watson integration would be able to tie a customer's question history to her record. If someone calls in after consulting Watson, the agent could pick up where the customer left off.
Gold says that 65 percent of calls to contact centers could have been resolved if the information could have been provided in a timely fashion. The implication is that those who use Watson could bring that number way down and customer satisfaction up. And with the Genesys integration going live this year, the results will be in soon.