Rather than treating automation as a narrow efficiency play, ClassPass approached the effort as a redesign of how support work is structured, combining automated resolution with real-time assistance for human agents and new roles focused on oversight and specialization.
Moving From Email to 24/7 Chat
Before introducing AI-driven tools, ClassPass relied primarily on email support, a model that became increasingly difficult to manage as its global customer base expanded. Agents worked across multiple systems, switching between knowledge bases, workflow tools and communication platforms to resolve issues.
“We had a growing consumer base and increasing CX costs,” said Sarah Vandenbrook, senior manager of CX operations and AI at ClassPass in a video for Decagon AI. “We were trying to find a solution that would allow us to continue to grow without increasing those costs at the same rate.”
Vandenbrook said the fragmented setup slowed response times and limited efficiency, particularly when agents were asked to handle more than one conversation at a time. To address that, ClassPass shifted support toward live chat and introduced AI-powered agent assistance to centralize documentation, workflows and response logic in a single interface.
“The most noticeable way this changed how our team works is that we are chat-first,” Vandenbrook said. By offering chat support 24 hours a day, seven days a week, the company moved the majority of agents who previously handled email into live chat.
She said real-time AI assistance improved agent confidence and speed, particularly when answering multiple chats simultaneously. Automated responses also began handling routine inquiries, reducing wait times for customers and limiting the number of issues requiring human escalation.
Redesigning CX Roles and Scaling Automation
As automation absorbed repetitive work, ClassPass reorganized its customer experience team. According to Vandenbrook, the company created dedicated quality-assurance roles focused on reviewing conversations and identifying where automated answers or workflows needed improvement.
“We’ve been able to create new roles,” she said, adding that team members moved into QA positions rather than frontline ticket handling.
ClassPass also introduced specialist roles for ticket categories that had historically generated higher friction or longer resolution times. That allowed experienced agents to focus on complex issues while the broader team handled general inquiries.
According to Decagon, ClassPass expanded chat support to 24/7 coverage, increased automated resolution significantly, and reported a 95% reduction in cost per support conversation, a figure cited by Decagon as a partner-reported outcome.
The company reported that automated ticket deflection was roughly 10 times higher than before deployment, as more routine issues were resolved without human involvement.
The Information reported that ClassPass’s support bot now handles a substantial share of inbound inquiries and, in many categories, performs on par with human agents. Vandenbrook said control over workflows was critical. “We wanted a tool where my CX team could manage any changes we wanted to make without needing to wait on someone else’s timeline,” she said, adding that updates to bot answers and custom workflows could be made immediately.
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.