Moderator
Cippy Seidler, Director, Consumer Care Center, Banner Health
Panelists
Heather Arthur, Vice President, Canada Contact Centres, Scotiabank
Jeff Grant, Senior Technology Consultant, Enterprise Contact Center, Southwest Airlines
Andy Lisk, Vice President, Global Head of Customer Service, StockX
At the 20th Annual Customer Contact West: A Frost & Sullivan Executive MindXchange, industry leaders discussed their unique use cases and the key stages of their AI journey—from initial planning and deployment to ongoing optimization. Discover what strategies proved effective, which pitfalls to avoid, and how AI is transforming operations in real-world scenarios.
Q&A HIGHLIGHTS
What is the most important AI lesson learned to date?
- We have a lot of lessons still to learn. The first data source was not current and was inconsistent. We had to choose a different data source, then found old processes and procedures coming up. We needed to clean up a lot of data.
- We started four years ago and just started using conversational AI this week. The data is very important. You have to make sure your articles are clean and clear. Take it slow with a couple of subtopics, about 7-8% of chat topics.
- Conversational AI: We made the decision to move to conversational AI. Before, 90% of conversations happened via email. We saw an uptick when we added chat bots and found customers preferred it. We wanted to use conversational AI to have consistent answers to customers and have agents work on more complex issues.
- Customer reaction: So far, the response has been positive for simple issues, but we have to keep training AI as processes and procedures change. Where customers are frustrated, the tech has to be tweaked.
- We are 2 years into a 5-year plan. There is a lot of work to do to clean up documentation and create policies to make them simple for chatbots and agents to use. We launched an analytic tool that reports most customer calls for banking information. Now can see how many calls are coming in and why they call. When we know the data, we can make data-driven decisions that will help make us more profitable.
How are agents responding to AI?
- AI is still a buzzword and can be scary for employees. If you can tell a story about automation, it can help them get connected to the future. For example, when ATM’s were introduced, it was scary for bank employees. But ATM’s just replaced functions that could be automated. The job descriptions of bank employees changed to be more advice-based.
- It’s helpful to ask employees to play with AI. Testing AI on the help desk helps employees to see if the tech is user-friendly, and where it might be improved. Get every person involved. There are benefits. Time saved will be invested in coaching and training to help employees work better, faster, and get more commissions.
- It’s important to point out to agents that AI can take over mundane tasks and free them to do more complex ones.
- Give the tech to agents first. Then get immediate feedback. When agents talk to others about the benefits AI, that builds enthusiasm for the process.
What is a good place to start?
- Email is the easiest place. Then take repetitive, mundane tasks from agents.
- We started small. We had a contact rate of 36% of orders. By putting Wizmo, where’s my order, into the flow, we reduced that number to 10%. It was a simple change with a big impact on the customer.
- Get the budget approved. Start today. Train agents to be digital adopters; to learn, process, and master AI automation tools. Example: Microsoft Copilot will summarize meeting notes, action plans, and add items to your calendar.
What is coming next for you in your specific area?
- Call wrap-up and summary. We’re looking at what we can provide to agents during customer interactions based on the conversation in real time. AI could remind the agent of newer processes, etc. It could bring in other teams not in the initial layout. It could extract data that is important or relevant.
- Digital voice. Having the ability to have a conversation with a customer via AI. We want to walk into that, making sure we have consistency, the right tone, and right language (not fake, “suspected inauthentic” language). AI will understand the tone of the customer, if the customer gets upset, the call will be transferred to an agent.
- We can’t do anything more until the next upgrade. Right now, recordings are not capturing, we have to go to the cloud. We will be going to workforce management in new ways. AI will catch and remember good calls and inform agents.
How are you justifying costs? Who did you have to meet with?
- We had to fix the contact rate. We meet weekly, review initiatives, and find anything off track that needs to be fixed.
- We looked at outside industries for comparisons. We are often compared to other banks, and the technology gap needed to close. We made a company investment and involved digital teams and others in the ask. There was a tech fail, and we took that opportunity to make the ask.
- Center for CX. Investing in customer experience. Trying to recover from the pandemic. Our field is leisure travel; what can we do to give the best experience on the phone to help the customer get to a special birthday trip or special occasion.
