Your meetings already have the answer
Tell us what mission you're trying to run – we'll help you
chart the flight plan.
MISSION
Customer sentiment analysis that sends happy and unhappy moments to Slack
Praise and problems both come up on customer calls, then get buried. This mission sends happy moments and escalations to Slack as they happen and writes a weekly sentiment read across your customer calls.
Inspired by
Team Fathom
Mission type
Customer Success & Enablement
Best for
Customer Success
Sales
Executive/Leadership
Output
Real-time happy and escalation alerts in Slack, plus a weekly sentiment report
System used
Fathom
What it does
This mission runs customer sentiment analysis on every customer call. Two Fathom Trackers listen for happy moments and for escalations, and each one posts the matching clip to its own Slack channel, so your team hears about praise and problems while they're still fresh.
Once a week, a skill in ChatGPT or Claude reads that week's customer call transcripts through Fathom and writes a sentiment report: what customers are happy about, what's frustrating them and which accounts need attention.
Who it's for
Customer success and sales leaders who want to hear about an unhappy customer before it turns into a churn conversation, and want the wins to reach the team too.
The flight plan
Set up the trackers once, then run the sentiment report every week. For accounts already at risk, pair it with the churn risk mission.
- Create two Trackers in Fathom: one for happy customer moments and one for customer escalations.
- Test each tracker's wording with Search with AI on past calls before you save it.
- Limit both trackers to your Sales and CS teams' customer calls.
- Subscribe a Slack channel to each tracker, like #customer-wins and #customer-escalations.
- Connect Fathom to ChatGPT or Claude and run the starter prompt each week for the sentiment report.
- If a tracker fires on small comments, rewrite it around overall sentiment instead of single words.
RULES
Keep happy and unhappy trackers separate.
Describe overall sentiment, not single words like "great" or "issue."
Send escalations to a channel someone owns and checks every day.
Quote the customer in every report item and link to the moment in Fathom.
Rank accounts by how serious the issue is, not how often it came up.
What good looks like
A good alert tells the team who's unhappy, why and what's at stake. A mediocre one flags every call where someone said "issue."
GOOD
"Escalation, Brightline Logistics QBR: 'We've raised the reporting bug three times and it's still blocking our exec dashboard.' Renewal in 60 days. Owner: Priya (CSM). Clip linked."
MEDIOCRE
"Negative sentiment detected in 4 calls this week."
STARTER PROMPT
⧉ CopyUse Fathom to find all of our customer calls from the past week and read the full transcripts. Write a customer sentiment report with three sections: what customers were happy about, what frustrated them and which accounts need attention this week.
For each item, quote the customer, name the account and link to the moment in the call. Rank the accounts that need attention by how serious the issue is and what's at stake, like an upcoming renewal. Skip small talk and passing comments.
FAQs
Questions? We’ve got you covered.
How do you run customer sentiment analysis on sales and CS calls?
Set up Fathom Trackers for happy moments and escalations, and subscribe a Slack channel to each so matching clips post as they happen. For a weekly view, connect Fathom to ChatGPT or Claude and have it read the week's transcripts and write a sentiment report.
Why does my sentiment tracker flag calls that aren't negative?
Trackers built on single words pick up small comments. Describe overall sentiment instead, like a customer expressing frustration with the product or service, and test the wording with Search with AI on past calls before you save it.
Which Fathom plan do I need for this mission?
Trackers and their Slack alerts are on the Team and Business plans. The weekly sentiment report runs on any Fathom plan, including Free, through Fathom's MCP integration with ChatGPT and Claude.
What is customer sentiment analysis?
Customer sentiment analysis measures how customers feel about your product or service from what they say and write. Calls are one of the richest sources, because customers say things there they never put in a survey.
How is this different from a churn risk alert?
A churn risk alert looks for signs that an account might leave. Sentiment analysis covers the whole picture, including what customers love, so your team hears about the wins as well as the problems.
New missions, straight to your inbox
A new mission launches every 1–2 weeks – real ways Fathom users are turning meetings into content, intelligence, and outcomes nobody expects. Subscribe and we'll send the new ones as they go live.
