Customer support · Production

Xanterra Travel Collection achieves 74% containment rate and $3.3M revenue lift with Cresta AI Agents

The problem

Xanterra's Central Reservations team faced extremely high inquiry volumes across multiple travel brands and four contact centers, making it difficult to maintain fast, consistent, and personal service at scale. QA was entirely manual and managers lacked the visibility and tools needed to coach agents effectively.

Workflow diagram · grounded in source
1
Guest inquiry via chat or voice
trigger
“roll out AI Agent for voice and chat across its Central Reservations team, starting with routine guest inquiries over chat”
2
AI Agent handles inquiry autonomously
ai_action
“"Faith," Yellowstone's AI Agent, is achieving a 76% containment rate, while "Cali," the AI Agent for Death Valley, is containing 62% of chats without agent intervention and "Skye", the AI Agent for Glacier National Park, is seeing a 84% …”
3
Insights feed into Agent Assist
integration
“Insights from AI Agent feed back into Agent Assist and Conversational Intelligence, giving live agents real-time guidance, automated summaries, and analytics that reinforce sales behaviors and continuously improve service quality.”
4
Agents use AI as learning resource
ai_action
“Agents have embraced AI Agent as a learning resource, using it to quickly find answers during live chats and improve their knowledge in real time.”
5
AI Analyst surfaces actionable insights
ai_action
“Xanterra adopted AI Analyst to turn contact center data into actionable insights. By surfacing trends across brands and providing supporting evidence, AI Analyst helps leaders make smarter decisions that directly impact both customer exp…”
6
Continuous optimization loop
feedback_loop
“Continuous performance analysis driving ongoing AI Agent optimization”
Reported outcome

Xanterra achieved a 74% containment rate across AI Agents and a $3.3M revenue increase through reinforced sales behaviors, while avoiding hundreds of thousands in potential guest recovery costs.

Reported metrics
Overall containment rate74%
Revenue lift$3.3M
containment rate — Faith (Yellowstone)76%
containment rate — Cali (Death Valley)62%
Show all 10 reported metrics
overall containment rate74%
revenue lift$3.3M
containment rate — Faith (Yellowstone)76%
containment rate — Cali (Death Valley)62%
containment rate — Skye (Glacier)84%
incremental revenue through reinforced sales behaviors$3.3M
agents on active coaching plansDozens
1:1 coaching sessions loggedhundreds of 1:1 sessions
guest recovery costs avoidedhundreds of thousands in potential guest recovery costs
AI Agents planned for Central Reservations9
Reported stack
CrestaAI AgentAgent AssistConversational IntelligenceAI Analyst
Source
https://cresta.ai/customers/xanterra
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Xanterra achieved a 74% containment rate across AI Agents and a $3.3M revenue increase through reinforced sales behaviors, while avoiding hundreds of thousands in potential guest recovery costs.

What tools did this team use?

Cresta, AI Agent, Agent Assist, Conversational Intelligence, AI Analyst.

What results were reported?

Overall containment rate: 74%; Revenue lift: $3.3M; containment rate — Faith (Yellowstone): 76%; containment rate — Cali (Death Valley): 62% (source-reported, not independently verified).

How is this customer support AI workflow structured?

Guest inquiry via chat or voice → AI Agent handles inquiry autonomously → Insights feed into Agent Assist → Agents use AI as learning resource → AI Analyst surfaces actionable insights → Continuous optimization loop.

Related customer support cases