Workshop: Why Negative Sentiment Doesn't Predict Churn
About Guest
Alex Atkins is the VP of Growth and Operations at Sturdy.ai, an AI-powered customer intelligence platform that turns customer conversations across emails, chats, tickets, and calls into actionable insights for B2B teams. Drawing on a background in psychology, communications, marketing, business analytics, and growth, Alex helps teams better understand churn signals, retention risks, and expansion opportunities so they can act faster and more intelligently.
In this Workshop
Customer churn rarely happens because of one major event. Instead, it's the result of multiple signals building over time. Many of those signals go unnoticed until it's too late.
In this workshop, Alex Gluz speaks with Alex Atkins, VP of Growth and Operations at Sturdy.ai, about what analyzing 47 million B2B customer conversations reveals about customer churn. They discuss why traditional customer health scores, NPS, and CRM data often fail to predict churn. The conversation also explores how AI-powered customer intelligence helps B2B companies identify churn risk, improve customer retention, and uncover new revenue growth opportunities.
Workshop Highlights:
01:00 – Meet Alex Atkins, VP of Growth and Ops at Sturdy
01:32 – Acquisition, retention, RevOps, and analytics: why they are all the same problem
04:31 – What Sturdy does: analyzing every customer conversation at scale
07:00 – The 47 million conversation data set
08:07 – What CRM fields and health scores will never show you
09:39 – Why negative sentiment does not meaningfully predict churn
10:40 – The contract request and executive change: an 80% churn predictor
12:02 – The over promise signal and why onboarding is the highest-risk stage
14:17 – Product changes, internal communication failures, and CS fallout
16:07 – Why product, CS, and leadership operate on different timelines with different information
19:26 – Why NPS and health scores miss what actually matters
23:44 – The gap between seeing a signal and acting on it
26:17 – Standardizing responses and building playbooks for every signal
29:16 – Where growth leaders should put their budgets right now
33:02 – Paid media and the expansion motion: using customer signals
35:33 – Generating case studies on demand with conversation intelligence
Resources Mentioned
Quotable Moments
"Churn is almost never caused by one specific thing. It's a combination of signals that matters most."
"The most dangerous finding in our report is that sentiment is a very weak leading indicator."
"Without that intelligence and timing, especially today, you're going to be falling behind."
"When you start to hit three of these different signals at one time, you know you're in deep trouble."
"If you don't know the why behind the what, then you really don't know anything."
Action Steps
- Use AI-powered customer intelligence: Analyze customer conversations in real time to identify churn risks before they become lost accounts.
- Create a single source of customer intelligence: Align sales, customer success, product, and leadership around the same customer signals and insights.
- Build playbooks for key customer signals: Standardize responses to events like executive changes, product issues, and contract requests so teams can act quickly.
- Look beyond customer health scores: Combine quantitative metrics with customer conversations to understand why accounts are at risk and where expansion opportunities exist.
- Turn customer insights into a growth strategy: Use conversation intelligence to improve customer retention, strengthen expansion efforts, and build higher-performing marketing campaigns.
