Customer Contact Week offers an excellent opportunity to learn from industry peers pushing the boundaries of CX innovation.
The customer support team at JetBlue is an excellent example, augmenting Amelia – its virtual agent – with generative AI (GenAI) and achieving positive results.
After three months, it secured a 12 percent containment boost, with a more modest – albeit positive – uplift in customer satisfaction (CSAT) and first contact resolution (FCR).
However, perhaps most critically, JetBlue’s human agents – or “crew members” – are onboard. That’s largely thanks to a careful strategy of empowerment and prepping agents for the future of AI, which started long before the introduction of Amelia.
“Empowerment isn’t just a buzzword; you have to show agents what empowerment looks like in practice,” said Shelly Griessel, VP of Customer Support at JetBlue.
We believe the role of a customer support person is evolving, and we need to ensure they’re equipped for it.
In this vein, Griessel shares several best practices for supporting agents in handling more complex tasks before offering advice for augmenting a high-performing team with AI.
Prepare Human Agents for Their Future Jobs
First, JetBlue ensures that crew members are set up to thrive in their jobs. There are several ways in which the company does so.
For instance, it caps occupancy rates at 85 percent and targets part-timers in recruitment, working a maximum of 30 hours a week to avoid burnout.
Additionally, every team leader sets aside two hours a week for crew members to engage with them about anything.
According to Griessel, these two-hour periods are always busy, giving agents an opportunity to discuss development opportunities, share concerns, and suggest ways to improve experiences.
There are also monthly “Direct Meetings” to review the good and the bad from across the company and promote that culture of open communication.
Griessel shared how this has helped facilitate an optimal team climate that drives a positive culture. She noted:
We focus heavily on culture, keeping close communication with crew members and making sure they understand the “why” behind everything, from AI to outsourcing.
Yet, when it comes to Amelia, JetBlue has gone far beyond communicating with crew members to ensure success; it’s also collaborating with them.
Cautiously & Collaboratively Introduce AI Agents
When JetBlue introduced the concept of AI Agents, crew members voiced their job security anxieties.
Griessel shared a strong message to overcome these: “We owe it to you to keep you relevant as the industry evolves.”
She then shared a vision for focusing on meaningful interactions rather than routine calls and of how AI will complement their current roles.
While the message of: “Amelia 2.0 will work the shifts no one else wants,” also helped, that vision of complementary AI struck a chord.
Soon, crew members started supporting the development of Amelia, so their virtual colleague took more of the customer contacts they didn’t want to handle.
In doing so, JetBlue’s team reviews automated interactions, guides improvements, minimizes the chances of hallucinations, and fast-tracks Amelia’s learning.
However, that’s skipping ahead a few steps, as Griessel offered more advice for successfully deploying a contact center AI Agent.
Action the Optimal Opportunities for GenAI
Before implementing an AI Agent, contact centers must gain a granular understanding of their demand drivers. JetBlue’s conversational AI provider – ASAPP – offered a helping hand here.
Griessel noted: “ASAPP has helped us greatly in understanding the intent behind customer calls, which allowed us to identify areas where we can improve, like policies on baggage.”
Indeed, JetBlue could prioritize its primary contact reasons, ensure the AI agent has the necessary knowledge to handle applicable queries, and orchestrate effective experiences.
Of course, not every contact reason is possible to automate, like refund decisions. These are out of Amelia’s scope due to regulatory scrutiny, so JetBlue and ASAPP have added guardrails to ensure such queries escalate immediately to a crew member.
Nevertheless, Griessel claimed: “With guidance from our top crew members, Amelia is learning incredibly quickly.”
Don’t Wait Around for the Perfect Moment to Introduce AI Agents
There are several reasons why a brand may postpone an AI Agent deployment, from technical debt to knowledge base limitations.
Yet, Griessel warned that no environment is ever “perfectly ready”. She said: “We plan, but things always change, and we don’t accomplish everything we set out to.”
I believe you have to jump in, even if the timing isn’t perfect, or you’ll be left behind – especially in terms of cost efficiency, but also customer satisfaction.
In making this point, Griessel stressed that the most valuable commodity for customers today is time, so speedy and consistent answers are crucial.
Without that consistency, they’ll keep calling back, and leaders will forever stay firefighting and not planning ahead.
“There’s no luxury in waiting until you’re “ready”,” concluded Griessel. “By putting boundaries around what the AI can handle, you limit the risks. So my advice is to start now and adapt as you go.”
For those looking for a conversational AI partner, ASAPP is one option. But there are also plenty of others. So, alongside ASAPP, check out the vendors in CX Today’s Marketplace.