Customer support · Production

Snap Finance cuts AHT 40% and raises containment 5.5x with Cresta generative AI

The problem

Snap Finance was experiencing rapid growth and struggled to scale its contact center while managing costs and compliance. Reporting systems were inadequate and quality management relied on random call sampling, giving supervisors little visibility into overall agent performance.

First attempt

Previous technology was described as anemic with major reporting problems, and quality management was effectively arbitrary—only a handful of random calls were reviewed each month.

Workflow diagram · grounded in source
1
Customer contacts via voice or chat
trigger
“Channel: Voice & Chat”
2
Virtual Agent resolves digital inquiries
ai_action
“With Cresta's Virtual Agent chatbot, the team was able to uncover conversation behaviors and drivers and resolve more of the common inquiries through digital channels on their website. "Pushing volume from voice into digital is not an ea…”
3
Agent Assist real-time prompts
ai_action
“With assists coming to the agents in real-time, they have seen average handle time (AHT) reduce by a staggering 40%”
4
100% QA monitoring of all calls
validation
“the team has overhauled its approach to quality management, and can monitor 100% of calls in real time, from both a quality and a compliance perspective”
5
Supervisor real-time coaching
feedback_loop
“Powered by Agent Assist, supervisors can coach agents in the moment to help with calls, ensuring that all agents, regardless of where they are geographically, are speaking "one cohesive language"”
Reported outcome

Cresta enabled Snap Finance to reduce average handle time by 40%, raise deflection rate from 6% to 33%, achieve 100% QA automation across all calls, and improve customer satisfaction by 23% along with employee engagement scores.

Reported metrics
Containment rate5.5x
CSAT23%
Average handle time reduction40%
Deflection/containment ratefrom 6% to 33%
Show all 7 reported metrics
containment rate5.5x
CSAT23%
average handle time reduction40%
deflection/containment ratefrom 6% to 33%
QA automation coverage100%
Employee Net Promoter Score and engagementincreased
agent compliance scorescompliance scores higher
Reported stack
CrestaVirtual AgentAgent Assist
Source
https://cresta.ai/customers/snap-finance
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Cresta enabled Snap Finance to reduce average handle time by 40%, raise deflection rate from 6% to 33%, achieve 100% QA automation across all calls, and improve customer satisfaction by 23% along with employee engagem…

What tools did this team use?

Cresta, Virtual Agent, Agent Assist.

What results were reported?

Containment rate: 5.5x; CSAT: 23%; Average handle time reduction: 40%; Deflection/containment rate: from 6% to 33% (source-reported, not independently verified).

What failed first in this deployment?

Previous technology was described as anemic with major reporting problems, and quality management was effectively arbitrary—only a handful of random calls were reviewed each month.

How is this customer support AI workflow structured?

Customer contacts via voice or chat → Virtual Agent resolves digital inquiries → Agent Assist real-time prompts → 100% QA monitoring of all calls → Supervisor real-time coaching.