Customer support · Production

Frame.io uses Intercom to provide personalized support to 2M customers

The problem

Frame.io's customer base grew more than 11 times in size, creating pressure to scale support beyond the founding team. Previous tools Help Scout and Zendesk each fell short: Help Scout's ticketing wasn't suitable for their needs, and Zendesk required significant design and engineering resources to configure.

First attempt

Help Scout's ticketing solution was not suitable for Frame.io's needs, and Zendesk required design and engineering resources to configure rather than being a plug-and-play solution.

Workflow diagram · grounded in source
1
Customer reaches out
trigger
“managing 4,500 conversations each month”
2
Self-serve help center
output
“We have over 20,000 help center visits and 36,000 article views every month”
3
Custom Bot handles FAQs
ai_action
“implemented a custom bot flow to help resolve some of their most frequently asked questions”
4
Auto-assign via workload management
routing
“automatically assigns inbound conversations to ensure that no agent has more than five conversations in their active queue”
5
Agent handles conversation
human_review
“the benefits of having their Product and Sales teams able to jump into conversations in the Inbox and collaborate to reach faster resolutions”
6
Proactive customer alerts
output
“leverages proactive support to alert customers when a temporary issue arises”
7
Conversation tag analysis
feedback_loop
“utilizing conversation tags to analyze recurring trends and identify where customers are running into issues”
Reported outcome

First-response time was reduced from 3 hours to 15 minutes (an 87.5% reduction), the team now supports two million customers handling 4,500 conversations per month, and workload management improved team morale.

Reported metrics
First-response time reduction87.5%
First-response time before3 hours
First-response time after15 minutes
Customers supportedtwo million
Show all 9 reported metrics
first-response time reduction87.5%
first-response time before3 hours
first-response time after15 minutes
customers supportedtwo million
monthly conversations managed4,500
monthly help center visitsover 20,000
monthly article views36,000
customer base growthmore than 11 times
team morale impactgreat impact on team morale
Reported stack
IntercomCustom BotsHelp ScoutZendesk
Source
https://www.intercom.com/customers/frame-io
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

First-response time was reduced from 3 hours to 15 minutes (an 87.5% reduction), the team now supports two million customers handling 4,500 conversations per month, and workload management improved team morale.

What tools did this team use?

Intercom, Custom Bots, Help Scout, Zendesk.

What results were reported?

First-response time reduction: 87.5%; First-response time before: 3 hours; First-response time after: 15 minutes; Customers supported: two million (source-reported, not independently verified).

What failed first in this deployment?

Help Scout's ticketing solution was not suitable for Frame.io's needs, and Zendesk required design and engineering resources to configure rather than being a plug-and-play solution.

How is this customer support AI workflow structured?

Customer reaches out → Self-serve help center → Custom Bot handles FAQs → Auto-assign via workload management → Agent handles conversation → Proactive customer alerts → Conversation tag analysis.

Related customer support cases