← All Work
Customer SuccessChurn ReductionOnboardingPlaybooks

Building Customer Success Systems That Cut Churn by 44%

A food subscription business was losing nearly a third of subscribers every month with no system for spotting at-risk customers. We built a customer success function from scratch: playbooks, tooling and measurement. Monthly churn dropped by 44%.

ClientA food business
IndustryFood & Beverage / Subscription
Year2023
Duration12 weeks
EngagementEmbedded Operator
ToolsIntercom, Notion, Google Sheets, WhatsApp Business API
The Outcome

By the numbers

Monthly Churn Rate
Before32%
After18%
−44%
Support Response Time
Before18 hrs avg
After2.4 hrs avg
−87%
D45 Retention
Before51%
After74%
+45%
NPS Score
Before22
After61
+39 pts
The Problem

What we walked into

At 32% monthly churn, the business was essentially on a treadmill, acquiring customers only to lose them before they became profitable. The support function was reactive, response times were slow, and there was no NPS measurement or early-warning system for customers about to cancel.

The Approach

How we did it

01 · Diagnosis

Understand who churns and why

We started with exit interviews, 40 calls with customers who had cancelled in the previous 60 days. The patterns were clear: poor onboarding experience, slow response to complaints, and no proactive outreach in the first two weeks. We also found that customers who received a personal check-in message by day 7 had 3× the D45 retention of those who didn't.

02 · Systems Build

Create the infrastructure

We built a customer health scoring model in Google Sheets, integrated with Intercom to flag at-risk customers. We wrote onboarding sequences, complaint resolution playbooks, and escalation protocols. Every interaction type had a documented response framework, so quality didn't depend on who picked up the ticket.

03 · Measurement

Track what matters

We set up NPS measurement at D14 and D30, response time tracking in Intercom, and churn cohort analysis in Google Sheets. The weekly CS review became a data session: what are the numbers saying, and what do we do about it? Within 12 weeks, churn had dropped from 32% to 18%, and NPS had jumped from 22 to 61.

The thing I appreciate most is that she speaks in data. Every recommendation had a measurement framework attached. That's rare in lifecycle marketing.

Co-founderFood Business
Work Together

Want results like these?

Let's talk about what's broken in your lifecycle and what it would take to fix it.