Handling Complex Calls via Conversational AI Voice Assistants

As we opened our hotels and casinos after the initial COVID shut down in 2020, we realized that the contact center as we knew it was about to change. While the demand quickly returned, the ability to find and recruit staffing became a serious challenge.

Not all of the previous employees were eager to return to the office. With increased demand and limited staffing, this had a major impact on our ability to answer the call volume and provide a good customer experience. We had to pivot from our normal way of doing business. We had to open more options to remote workers and figure out how to handle the increase in calls with limited staffing.

Fortunately, we had been exploring remote working capabilities prior to COVID on a limited emergency and disaster recovery basis. We were able to expand those processes to give more agents the opportunity to work remotely. Using the WFM and QA software we had in place, we were able to ensure that the staff we did have were as productive and efficient as possible and continued to deliver a great experience.

Long wait times and high abandon rates were issues that we faced. Due to the challenges with recruiting a staff, we had hold times ranging from 45 minutes to an hour and were abandoning up to 50 percent of the calls. I shared with my CFO the potential loss of revenue due to the abandoned calls and that we needed to find another way to handle these calls. We could no longer just add bodies to answer the phones.

"We built a conversational AI voice assistant to handle these calls in 4 weeks. I hoped that it would handle about 40 to 50 percent of these calls without going to an agent. I was extremely pleased that from day one it handled 87 percent of the calls."

I had been looking into artificial intelligence (AI) and began speaking with several AI vendors about solutions. I was skeptical of AI because of previous negative experiences. I did not want my customers to speak to a robot. I did not want to provide a bad experience for my customers, but I also knew that not answering the phone was a worse experience.

I was able to find a vendor with an exceptional conversational AI voice assistant. I gave them call samples of my best agent, and they developed a voice that sounded like my best agent's. We began by breaking down call types to identify a simple call type to handle initially that would give me the quickest relief. We started with simple transfer calls from the hotels, such as can you transfer me to housekeeping, front desk, spa, etc. We also added 20 FAQs that could be answered quickly and efficiently, such as "What time is check-in?" Do you allow pets? Do you have free parking?

The use case we identified was about 40,000 calls a month that were very simple and quick. We built a conversational AI voice assistant to handle these calls in 4 weeks. I hoped that it would handle about 40 to 50 percent of these calls without going to an agent. I was extremely pleased that from day one, it handled 87 percent of the calls.

This freed up my agents to be able to handle more complex calls and drive revenue. It immediately reduced my abandons and increased my revenue. We are now expanding to other use cases to become even more efficient and deliver a better customer experience.