Back to Use Cases

Cost Reduction Through AI-First Support

ClassPass experienced increasing interaction volume with high operational costs. Cyclical demand made staffing unpredictable, and their tech stack lacked automation integrations across channels.

📌Key Takeaways

  • 1Cost Reduction Through AI-First Support addresses: ClassPass experienced increasing interaction volume with high operational costs. Cyclical demand mad...
  • 2Implementation involves 5 key steps.
  • 3Expected outcomes include Cost Reduction: 95% per ticket.
  • 4Recommended tools: decagon.

The Problem

ClassPass experienced increasing interaction volume with high operational costs. Cyclical demand made staffing unpredictable, and their tech stack lacked automation integrations across channels.

The Solution

After evaluating 12 AI solutions, ClassPass chose Decagon. They deployed a generative chatbot plus Agent Assist integrated into Zendesk, expanding from 16 hours/5 days to 24/7 coverage.

Implementation Steps

1

Assess Current Challenges

ClassPass experienced increasing interaction volume with high operational costs. Cyclical demand made staffing unpredictable, and their tech stack lacked automation integrations across channels.

Pro Tips:

  • Document existing pain points and their business impact
  • Identify key metrics to track improvement
  • Map current workflows that need automation
2

Design the AI Solution

Plan the implementation of cost reduction through ai-first support using Decagon capabilities.

Pro Tips:

  • Configure Decagon for your specific requirements
  • Define success criteria and KPIs upfront
  • Identify integration points with existing systems
3

Implement and Configure

After evaluating 12 AI solutions, ClassPass chose Decagon. They deployed a generative chatbot plus Agent Assist integrated into Zendesk, expanding from 16 hours/5 days to 24/7 coverage.

Pro Tips:

  • Start with a pilot deployment on a subset of workflows
  • Test thoroughly with real-world scenarios
  • Train team members on the new system
4

Monitor and Optimize

Track performance metrics, gather feedback, and iterate on the configuration to improve outcomes.

Pro Tips:

  • Review performance dashboards weekly
  • Collect qualitative feedback from end users
  • Adjust thresholds and rules based on real data
5

Scale Across the Organization

95% cost reduction per reservation and 10x higher deflection than anticipated. Chat expanded from handling 5% to 80% of all support tickets.

Pro Tips:

  • Document best practices from the pilot phase
  • Create onboarding materials for new teams
  • Set up regular review cadences to maintain quality

Expected Results

Cost Reduction

3 months

95% per ticket

Deflection Rate

Immediate

10x higher

Coverage

Immediate

24/7 support

Channel Shift

6 months

5% → 80% chat

ROI & Benchmarks

Typical ROI

400-800%

within 6-12 months

Time Savings

70-90%

reduction in manual work

Payback Period

2-4 months

average time to ROI

Cost Savings

$200K-$1M+ annually depending on volume

Output Increase

10x ticket deflection improvement

Implementation Complexity

Technical Requirements

Medium4-8 weeks typical timeline

Prerequisites:

  • Existing helpdesk platform
  • Knowledge base content
  • Support workflow documentation

Change Management

Medium

Moderate adjustment required. Plan for team training and process updates.

Recommended Tools

Frequently Asked Questions

ClassPass experienced increasing interaction volume with high operational costs. Cyclical demand made staffing unpredictable, and their tech stack lacked automation integrations across channels.
After evaluating 12 AI solutions, ClassPass chose Decagon. They deployed a generative chatbot plus Agent Assist integrated into Zendesk, expanding from 16 hours/5 days to 24/7 coverage.
95% cost reduction per reservation and 10x higher deflection than anticipated. Chat expanded from handling 5% to 80% of all support tickets.
This use case is particularly relevant for Health & Wellness organizations, though the principles apply broadly to any team facing similar challenges.
Getting started requires Decagon access and integration with your existing systems. Most deployments begin with a pilot phase to validate results before scaling.

Last updated: February 2, 2026

Ask AI