Customer support · Production

Jukebox delivers instant buyer support at scale with Fin AI Agent, achieving up to 65% resolution rate and 40% conversion growth

The problem

As Jukebox scaled with 80% more orders and 83% growth in support volume, their lean team was overwhelmed by thousands of urgent monthly conversations and could not keep pace with buyer expectations for fast support — especially during peak-season queues that could hit thousands overnight.

First attempt

Their legacy platform Zendesk lacked live chat functionality, felt robotic and disconnected from their brand, and could not be customized to match Jukebox's pace or standards.

Workflow diagram · grounded in source
1
Knowledge base training
ai_action
“Jukebox started by training Fin on their entire knowledge base and product catalog. Giving Fin access to live product pages means it can guide buyers through those details in real time, helping them avoid mistakes and feel confident plac…”
2
Nuance instructions configured
ai_action
“The Jukebox team provided Fin with natural language instructions to shape how it handles nuanced scenarios, applies policy, and stays aligned with their brand voice and standards.”
3
Backend API integration
integration
“Jukebox connected Fin to their custom backend via API, giving it real-time access to order details, shipping status, production progress, and pricing.”
4
Buyer submits support query
trigger
“Orders can come in at any hour, and a buyer checking delivery times at midnight expects answers right away.”
5
Fin resolves query autonomously
ai_action
“Fin responds to questions about order details, delivery times, or cancellation, and the level of detail it gives makes people think they're talking to a person. It even handles back-and-forth.”
6
Urgent issues routed to human
routing
“If a message includes urgent terms like "not centered," "reprint," or "damaged" along with a file attachment, Fin was guided to instantly route the conversation to a human.”
7
AI insights drive improvements
feedback_loop
“Using AI powered insights, the team has recently identified a product that was consistently driving frustration and made the decision to remove it from their offering entirely.”
Reported outcome

Fin now handles up to 65% of support queries autonomously, covers 90% of peak-season queries, contributed to 40% conversion growth, and absorbed 83% yearly growth in conversations without adding headcount.

Reported metrics
Fin involvement rate50%
Fin resolution rateup to 65%
Peak-season sales queries handled by Fin90%
Conversion growth attributed to Fin40%
Show all 8 reported metrics
Fin involvement rate50%
Fin resolution rateup to 65%
Peak-season sales queries handled by Fin90%
Conversion growth attributed to Fin40%
Yearly growth in support conversations83%
Order volume increase80%
Cancel-and-repurchase rate within two minutes98%
Agent headcount avoidedsix or seven more agents
Reported stack
FinIntercom SuiteZendeskAI powered insights
Source
https://www.intercom.com/customers/jukebox
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Fin now handles up to 65% of support queries autonomously, covers 90% of peak-season queries, contributed to 40% conversion growth, and absorbed 83% yearly growth in conversations without adding headcount.

What tools did this team use?

Fin, Intercom Suite, Zendesk, AI powered insights.

What results were reported?

Fin involvement rate: 50%; Fin resolution rate: up to 65%; Peak-season sales queries handled by Fin: 90%; Conversion growth attributed to Fin: 40% (source-reported, not independently verified).

What failed first in this deployment?

Their legacy platform Zendesk lacked live chat functionality, felt robotic and disconnected from their brand, and could not be customized to match Jukebox's pace or standards.

How is this customer support AI workflow structured?

Knowledge base training → Nuance instructions configured → Backend API integration → Buyer submits support query → Fin resolves query autonomously → Urgent issues routed to human → AI insights drive improvements.

Related customer support cases