AI is no longer a nice-to-have but a business imperative. 70% of companies plan to increase resources such as budget, employee time, training, and technological infrastructure towards AI investments in 2024, according to a recent Slalom survey. The technology has strong applications for customer support across all industries. Recently, Slalom and Virgin Voyages teamed up to launch one of the first generative AI-powered digital humans built on Salesforce, named Vivi. The new chatbot underpins Virgin Voyages’ online customer experience with the goal of redefining customer satisfaction by creating a more engaging and realistic customer service chat experience.
I spoke with Katie Dunlap, general manager of global Salesforce at Slalom, a global business and technology consulting company, about Vivi and how other companies can implement generative AI into their business.
Gary Drenik: Slalom worked with Virgin Voyages to launch its new generative AI-powered chatbot, Vivi. Are there any examples of how Vivi has already been making a difference with Virgin Voyages and their Sailors?
Katie Dunlap:: The goal of implementing Vivi was to make it easier for Virgin Voyages’ Sailors to find the information they need with seamless and natural answers to inquiries. For example, imagine a question like, “I want to know if I can bring my service dog on the cruise, but I can’t find the right policies that showcase the differences between pets and service animals. Can you help?”
- A traditional chatbot could answer, “Here’s a link to our Service Animal Policy”
- With Vivi, the answer can be more human-like, “Hi Sailor, I’m sorry to hear you haven’t been able to find an answer to your question. Yes, your service dog is permitted on-board. Service dogs are defined as dogs that are individually trained to do work or perform tasks for the benefit of an individual with a disability, such as guiding people who are blind, alerting people who are deaf, pulling wheelchairs, etc. The task(s) performed by the dog must be directly related to the person’s disability.”
A recent Prosper Insights & Analytics survey highlights that nearly 82% of people prefer to communicate with a live person over an AI chat program when they need assistance with travel. Millennial and Gen-Z travelers are slightly more inclined to prefer AI chat programs compared to other generations. But answering the same questions over and over tends to burn out support staff, which is a huge challenge for the industry. A survey of 72 global customer contact enterprises found that retaining agents is the second highest priority for contact center leaders, with agent burnout highlighted as a major obstacle in meeting agent retention goals.
The question is, how can we influence customer behavior and encourage them to use an AI chat program for more simple questions? Vivi could potentially shift this behavior and is already easing workloads for Virgin Voyage’s Sailor Services team. In our generative AI-powered use cases, customers went to an agent 18% less, resulting in more budget, more time, and people doing jobs they want to do. We expect increased agent retention as a result since they can avoid answering the same questions all day. Plus, with the AI integration, customers avoid long wait times and get fast, easy answers. It’s a win-win for both sides.
Drenik: Now that Vivi is operational, where does she go from here? What could a future evolution of Vivi look like for Virgin Voyages?
Dunlap: We’re just scratching the surface of what’s possible with Vivi. We’ve been talking about integrating personalized use cases, such as booking itineraries, adding drink packages, and making dinner reservations. Today, an agent must get involved to book shore excursions and make booking updates. Wouldn’t it be great if Vivi turned a customer’s question into a sales opportunity? Say a Sailor is asking Vivi if they can use Sailor Loot to pay for drinks at one of the cruise destinations, it’s the perfect opportunity for Vivi to ask if they would like to purchase Sailor Loot.
It all comes down to which use cases will have the greatest impact on overall sales and agent workload. If a question is asked only a few times per month, it’s better to hand it off to an agent, but if it’s something that is asked often and Vivi could help increase the close rate, then it’s worth the investment.
Generative AI has made a lot of this possible. Before you had to manually enter questions and answers, but just think about all the ways you can ask a question and the time needed to develop these. With generative AI, the bot understands the question and the answer and can even address multi-staged questions. It requires less work to create an engaging bot and we’re getting better results – the best of both worlds.
In the future, as partners to brands like Virgin Voyages, we’re dreaming bigger. We’re talking about bringing Vivi on the ship with facial recognition to help with embarkation and disembarkation, dinner reservations, and ship activities.
Drenik: What questions do you get most often from companies looking to implement generative AI?
Dunlap: Most often companies are curious about data privacy and how to navigate compliance in a generative world. Since generative AI is a newer technology, there’s a misconception that company data is consumed by the large language model and shared with the planet, but we know that’s not how it works. Your data is owned by you and stays secure.
On the compliance front, there are many ways to leverage generative AI to meet your company’s branding and compliance guidelines. One of the greatest aspects of generative AI is that it will create answers for you from what it knows, and you can teach it how to answer right. However, in a world of compliance and branding, companies want a branded, well formatted response with certain wording for legal purposes. In this case, we’ve used generative AI to understand the question and guide to the right answer, but not generate the answer. It picks from a pre-built list of responses. This is a reasonable place to start that is still advantageous over current chat bots and website FAQs. For companies that have more flexibility, we can give generative AI more creative freedom to develop human-like responses.
Drenik: What’s unique about Slalom’s approach to generative AI?
Dunlap: We are pioneers, and the speed of our talented teams is a key differentiator. Over the last nine months we’ve built, tested, and delivered a framework to help companies implement generative AI. Through immersive tours and meetings, we’ve educated hundreds of companies on generative AI and possibilities for their business. By starting early, learning early, and creating our path, we’ve built a strong understanding of generative AI. Implementing generative AI can be daunting, due to regulations, branding concerns, and the rapid pace of technology, but the right partner can help address challenges, implement a customized solution, and ideate future iterations of the technology.
Drenik: Thanks Katie, for your insights on generative AI as it relates to the travel industry and beyond. It sounds like 2024 will be an exciting time for customer support as generative AI transforms the traditional chat bot. Whether you’re in travel, retail or healthcare, all industries can benefit from experimenting with AI and fine tuning the program for your needs.
Read the full article here
