COHORT PROGRAM — $950/SEAT · 3 WEEKS

Churn Analysis & Prevention

You’ve tried success calls, win-back emails, and feature releases. Churn keeps happening. In 3 weeks, you’ll diagnose which type of churn you have, build an early warning system from your product data, and design the intervention.

Jake McMahon, ProductQuant

Join the waitlist → See the curriculum ↓

For founders and product leaders at B2B SaaS with $500K–$10M ARR.

WHAT’S INCLUDED

Live Sessions 6 sessions, 75 min each, 2×/week

Churn Diagnosis Which of 6 archetypes you have, and in what mix

Early Warning System In-product churn signal dashboard built in your own tool

Intervention Map Which trigger fires at which signal, for each archetype

Win-Back Sequence Day 0–Day 89 email + in-app logic designed

$950/seat · cohort-based · limited seats

The churn problem this cohort is designed to solve

You know churn is happening. You don’t know why.

Exit surveys say one thing. Win-back calls say another. Your product data says something else, but you’re not sure how to read it. Without a clear diagnosis, every intervention is a guess.

You’re treating every churned customer the same way.

One-size-fits-all win-back emails. Generic success calls. But churn from a budget cut looks nothing like churn from value not realised. The intervention that works for one type makes zero difference for another.

You find out a customer is churning when the email arrives.

By then you have 30 seconds to change their mind. The signal was visible in the product data 60 days ago — login frequency dropped, a key feature went dark, session depth declined. Nobody was watching.

WHO THIS IS FOR

Built for founders who suspect the answer is in their product data.

Right fit

Founders or product leaders at B2B SaaS with $500K–$10M ARR
You have visible, recurring churn you can’t fully explain
You have event data in PostHog, Amplitude, or Mixpanel (even if the setup is messy)
You’ve tried surface interventions (NPS, success calls) and churn hasn’t moved
You can commit ~5 hrs/week for 3 weeks

Not the right fit

Pre-revenue or pre-churn — you need at least 6 months of customer data
You have no event tracking set up at all — start with the analytics audit first
Your churn is entirely from pricing or market timing, not product behaviour
You want someone to do the analysis for you — this cohort teaches you to do it yourself

THE FRAMEWORK

The 6 churn archetypes. Most products have 3 or more in their mix.

ARCHETYPE 01

Value Not Realised

The user signed up, got through onboarding, and never reached the outcome they came for. Most common cause: the path to value is too long or too unclear.

ARCHETYPE 02

Wrong-fit customer

The customer was never a good fit. They bought based on marketing or sales claims that didn’t match what the product actually does. Acquisition problem masquerading as retention problem.

ARCHETYPE 03

Competitor Switch

A specific competitor closed a feature gap or pricing gap that was the reason the customer stayed. Diagnosable from the timing of churns relative to competitor releases.

ARCHETYPE 04

Budget Cut

The customer churned because of economic pressure at their company, not product dissatisfaction. You can’t fix what isn’t yours to fix — but you can identify these fast and stop chasing them.

ARCHETYPE 05

Champion Left

The internal advocate who drove the purchase and renewal left the company. The product didn’t expand to other stakeholders before the champion left. Preventable with the right expansion motion.

ARCHETYPE 06

Feature Gap

The customer was waiting for a specific capability and it didn’t ship in time. Diagnosable from in-app searches, support requests, and usage drop-off patterns around the gap.

Week 1 of the cohort teaches you to diagnose which archetypes are dominant in your cohort — and in what mix. The intervention for each one is different.

WHAT YOU’LL LEAVE WITH

Five deliverables, all built on your own product and data.

Churn diagnosis report for your own product

Which of the 6 churn archetypes are dominant in your customer cohort, and in what proportion. Based on your product data, not your intuition.

In-product churn signal dashboard

The 3 in-product behaviours that predict churn 60 days out, built into a live dashboard in your own analytics tool. Your team sees at-risk accounts before they cancel.

Intervention map

Which Chameleon or Intercom trigger fires at which churn signal, for each archetype. Not a generic playbook — your specific trigger logic, mapped to your archetypes.

Win-back sequence design (Day 0–Day 89)

Email and in-app logic for each day of the win-back window. What fires when, what the goal of each touchpoint is, and how to measure if it’s working.

One churn experiment running live by end of Week 3

Not designed in theory — an intervention designed, configured, and live in your product before the final session. The cohort reviews results together.

THE CURRICULUM

3 weeks from diagnosis to live intervention.

WEEK 1

Diagnosing Churn: What Type Do You Have?

SESSION 1

The 6 churn archetypes

Value-not-realised, wrong-ICP, competitor-switch, budget-cut, champion-left, feature-gap. What each one looks like in your cohort data, how to tell which is dominant, and why the intervention for one type makes things worse if applied to another.

SESSION 2

Building the churn diagnosis from product data

Cohort analysis, feature adoption before churn, session patterns, and last-active signals — reading each one together using real product data shared by cohort participants. By the end of Session 2, you have a working hypothesis about which archetypes are dominant in your product.

Async — before Week 2

Run the churn archetype diagnosis on your own data using the worksheet provided. Identify your top 2 dominant archetypes with evidence from your product data. Share before Session 3.

WEEK 2

Building the Early Warning System

SESSION 3

Defining your churn prediction events

The specific in-app signals that predict cancellation 30–60 days out. Not NPS, not support tickets — specific usage patterns like login frequency drops, key feature inactivity, and session depth decline. How to identify which signals are predictive in your product specifically.

SESSION 4

Building the churn signal dashboard

Live configuration walkthrough in PostHog, Amplitude, and Mixpanel. You build the dashboard during the session. By the end, you have a live view of at-risk accounts segmented by signal strength.

Async — before Week 3

Finalize your churn signal dashboard. Identify the top 3 predictive behaviours for your product and the threshold at which each becomes an at-risk signal. Share in Slack for cohort review.

WEEK 3

The Intervention System

SESSION 5

Designing the intervention

In-app interventions (Chameleon, Intercom), email (Customer.io), and success motion for each churn archetype. What triggers what, when it fires, and what the success metric for each intervention is. How to avoid the mistake of applying the same intervention to every churn signal.

SESSION 6

The win-back sequence

Day 0 to Day 89 logic. What emails fire, what in-app experiences activate, what the goal of each touchpoint is, and how to measure if the sequence is working vs. just generating noise.

Async — before Session 6

Design your intervention map: which trigger fires at which churn signal, for which archetype. Submit for Jake’s review before the final session. Session 6 opens with a walkthrough of the cohort’s designs.

THE WORK

Real outcomes from the approach behind this cohort.

HEALTHCARE SAAS — RETENTION SYSTEM

Chameleon in-app screens across 6 retention branches

Customer.io win-back emails across 90 days

Full churn intervention system built: early warning dashboard, archetype-specific in-app flows, and 90-day win-back sequence. Each branch triggered by a different churn signal.

HR PLATFORM — CHURN ANALYSIS

95.8%

of 1M+ members never converted to paid

churn drivers identified from product data

$250K–$400K conversion opportunity identified from the churn analysis. Six distinct drivers found, each requiring a different intervention.

FORMAT

Live, small-cohort, applied to your real churn data.

Duration

3 weeks

6 live sessions, 2×/week, 75 minutes each

Async work

~2 hrs/week

Worksheets and builds applied to your own product data, reviewed before the next session

Cohort size

12 seats max

Small enough for real review. Jake looks at every archetype diagnosis and intervention map personally.

Platform

Zoom + Slack

Live sessions on Zoom. Async work and cohort discussion in a private Slack channel.

Timezone

GMT-friendly

Sessions scheduled to work for EU and US East. Exact times confirmed with cohort on registration.

Recording

All sessions

Every session recorded. Access for 12 months after the cohort ends.

WHO’S TEACHING

Jake McMahon — ProductQuant

Jake McMahon

8+ years building growth systems inside B2B SaaS · Behavioural Psychology + Big Data (Masters)

The churn archetype framework in Week 1 is the same one I use when I open a churn engagement with a client. The 27 in-app retention screens, the 90-day win-back sequence, the early warning dashboard — these came from real products, not from theory.

The cohort format exists because diagnosing churn is harder to learn from a video than from watching someone else work through it on their real data. When another participant runs their diagnosis in the session, you learn something you wouldn’t learn from a case study.

Sessions are capped at 12. I review every async submission personally. If your churn archetype diagnosis is wrong, I’ll tell you specifically why — with evidence from your data.

Questions.

Or get in touch →

What analytics tools does this work with? +

PostHog, Amplitude, and Mixpanel for the churn signal dashboard. Chameleon or Intercom for in-app interventions. Customer.io for email. If you’re on different tools, get in touch — the approach applies, the configuration walkthrough may differ.

How much historical data do I need? +

At least 6 months of customer history to run a meaningful churn diagnosis. Cohort analysis needs enough churned accounts to show a pattern — generally 20–30 churned accounts minimum. If you’re below this, the tools still apply but the statistical confidence will be lower.

Do I need Chameleon or Customer.io already set up? +

No. Week 3 covers the design of the intervention, not necessarily the live configuration. You can implement the intervention map in whatever tools you have. The logic transfers regardless of tooling.

When is the next cohort? +

Dates are confirmed when enough waitlist registrations are in. Join the waitlist for first access and no obligation to enroll.

Is this related to the Analytics Audit offer? +

The Analytics Audit is a done-for-you service; this cohort is a learning program where you build everything yourself. The audit is a good precursor if your event tracking is very messy, but it’s not a prerequisite for this cohort.