What does it really take to bring AI customer service to life inside a fast-moving, customer-obsessed ecommerce business?

At IPSY, it started with a bold hire and an even bolder vision.

When Anna Prince joined the IPSY as AI & Automation Strategy Manager in April, the Customer Care team was at a crossroads. An aging scripted bot and manual workflows were slowing down support. Routine issues consumed agent time. And while the desire for change was there, the path forward wasn’t clear.

It was time for something better. So Anna set out to reimagine how IPSY showed up for its millions of beauty-loving members.

Six months later, the change was dramatic. CSAT was up 41% and containment had reached 72%. IPSY’s AI agent, Glam Bot, was handling thousands of conversations end-to-end.

IPSY didn’t just automate customer service. They redefined what was possible and set a new bar for AI success across the company. Here’s how.

From scripted to smart: Why IPSY replaced its bot with an AI customer service agent

Before May 2024, IPSY’s customer support relied on a scripted automation system. It followed rigid decision trees, offering limited help when members brought up questions outside of predefined flows. It was predictable but not adaptable.

That was the tipping point. The Customer Care team knew their support experience needed to evolve. And with Anna in place, they had the leadership to drive it.

She quickly partnered with IPSY’s Internal Apps team to map a smarter path forward—one powered by generative AI. The vision was clear: build an AI agent powered by conversational AI for customer service that could understand natural language, personalize responses using member data, and resolve inquiries without requiring human handoff.

In May, IPSY launched its upgraded AI agent with Ada. Within weeks, customer satisfaction was climbing, and resolution rates were rising—a sign that this was more than just a tech change. It was a turning point in how IPSY delivered AI customer service.

IPSY’s strategy for big impact automation? High volume, low complexity.

Anna’s approach was pragmatic: start where automation would have the biggest impact, without sacrificing experience.

She zeroed in on missing and damaged item reports. These were high-volume tickets with straightforward workflows and low complexity. By connecting the AI agent to backend systems, IPSY gave it the power to validate requests and issue replacements—all without human involvement.

This wasn’t just a win for efficiency:

• The CSAT for missing items rose to 4.3, the highest of any support topic
• Containment steadily increased, freeing up agent time
• Resolutions were faster and more consistent for members

By September, thousands of member issues had been resolved entirely by the AI agent—quickly, accurately, and without escalations. The team proved that AI could deliver real value for the business and for the customer.

Can conversational AI also handle complex customer inquiries? IPSY proves it can

Next, Anna turned to a more complex use case: cancellation and retention.

Previously, when members told Glam Bot they wanted to cancel, the system didn’t offer any alternatives. The experience ended there—no save attempt, no personalization, no support.

Anna and the team changed that. They rolled out a retention playbook that empowered the AI customer service agent to have a conversation, not just a transaction. Smarter and more member-centric, their AI customer service agent could help people stay for the right reasons, not just delay the inevitable.

What business results can AI customer service deliver, and how does it impact your team?

In just four months, IPSY’s Customer Care team delivered a transformation that reached far beyond support tickets. From May to September:

• CSAT increased from 2.63 to 3.70, a 41% improvement
• Automated Resolution rose by 64%
• The team achieved a 943% ROI, with the AI agent resolving approximately 160,000 conversations end-to-end

These weren’t surface-level wins. They represented a deeper operational shift—one where customers received faster, more personalized service, and agents could focus their time on nuanced, high-impact work.

Today, IPSY’s AI customer service agent is the largest deployment of its kind inside the company. The success has captured the attention of other teams—from operations to marketing—who are now exploring how AI can unlock new opportunities in their own domains.

And Glambot? IPSY plans to:

• Expand into new channels like email to support a more seamless, omnichannel customer service experience
• Deepen backend integrations to enable full end-to-end workflows
• Explore AI-driven commerce use cases to offer personalized product recommendations in real time

And through it all, IPSY has invested in growth, not cuts. Anna Prince now leads the program full-time, and the team continues to grow. What started as a support initiative has become a company-wide catalyst.

The future of AI customer service is already here

IPSY didn’t just implement AI. They made it a competitive advantage. By empowering a strong leader, investing in the right platform, and focusing on real outcomes, the Customer Care team turned support into a launchpad for innovation.

Their story is proof that AI customer service isn’t about replacing people or chasing trends. It’s about giving teams the tools to do their best work and creating experiences customers actually enjoy.

What started as a support upgrade has become a blueprint for what’s possible when AI meets initiative.