solving retention · does the bucket hold water?

This essay is part of the Solving Distribution chain framework. If you haven’t read the main pillar article, start there to see how this link sits downstream of Validation and upstream of everything else.

The Distribution Chain: Link 1 of 7

Once you have validated that real people have a problem and will pay to solve it (Link 0), the single most expensive mistake you can make is to immediately try to grow.

The instinct is almost irresistible. We have our first paying users! Time to launch on Product Hunt, run ads, scale the outreach.

Resist it.

Before you spend a single euro or a single weekend pouring new traffic into your product, you have to answer one brutal, foundational question: do the people who join actually stick around?

This is Link 1 of the Distribution Chain: Retention.

If distribution is a bucket and acquisition is the water you pour into it, retention is the structural integrity of the bucket itself. Scaling before you solve retention accelerates burn. You spend money to acquire users who leak straight out of the bottom, and you are left with a rising churn rate, a damaged reputation, and empty pockets.

The Math of the Leaky Bucket

To understand why retention gates acquisition, you have to see how churn compounds.

Adjust the sliders above and watch what a few points of churn do. At 200 sign-ups a month, 3% monthly churn supports a business of roughly 6,700 users. The same 200 sign-ups at 10% churn support 2,000. Same marketing, same spend, a ceiling three times lower.

That ceiling is the whole point:

Steady sign-ups divided by monthly churn is your ceiling. 200 a month at 5% churn is a 4,000-user business. Forever.

A sticky bucket does more than keep users. It preserves your margin, so every new customer adds compounding revenue instead of replacing revenue that leaked out last month.

What Retention Actually Answers

Retention measures whether your product creates recurring, long-term value or a temporary pulse of novelty. It answers two questions:

• The Core Value Question: does this product solve the problem so well that losing it would actively disrupt the user’s workflow or life?
• The Habit Question: does using the product integrate naturally into a routine, or does it require continuous external prompting to remember it exists?

Real retention is a product output. It happens when the product keeps delivering on its original promise, not when you set up aggressive re-engagement sequences or wave a discount code at the cancel screen.

How to Diagnose a Broken Link 1

A retention failure is the silent killer in software products, because early revenue can mask it entirely while you are acquiring new users faster than you lose the old ones.

Quantitative Signals

• Downward-sloping cohort curves: plot each monthly cohort over Day 1, Day 30, Month 3, Month 6. A healthy curve drops and then flattens out parallel to the x-axis. A broken one keeps sloping toward zero.
• High net revenue churn: you lose more monthly revenue from churned customers than you gain from expansions and upgrades inside your existing base.
• Short customer lifetime: the average subscriber leaves after one or two billing cycles, which caps LTV low enough that no acquisition channel can ever pay back.

Qualitative Signals

• “Nice to have” feedback: ask churned users why they left and almost nobody says the product broke. They say “we just stopped using it” or “it was cool, but we didn’t have time for it”.
• Low PMF score: fewer than 40% of active users say they would be “very disappointed” if your product disappeared tomorrow, which is Sean Ellis’s product-market fit threshold.
• Usage ghosting: the user pays, logs in once or twice in week one, and never appears again until the renewal charge triggers a refund request.

Why Retention Breaks (Root Causes)

Why do users churn even after paying you money? It usually comes down to three structural breakdowns.

1. The Value Horizon Gap

Your product solves the problem, but it solves it once. An audit tool that hands a user a list of fixes on Day 1 delivers enormous value immediately. If there is no reason to run that audit again next week, a monthly subscription starts to feel wasteful, and the customer cancelling is the customer correcting your pricing model.

2. No Internal Trigger Loop

Behavioural design runs on triggers. If a user only remembers your product when your automated marketing email lands, that is an external trigger and your retention is fragile. High-retention products anchor to a trigger the user already has: the inbox overflowing, a new branch being cut, Friday payroll coming round again.

3. A Misaligned Ideal Customer Profile

Sometimes churn has nothing to do with product quality. If you sell an enterprise-grade tool to cash-strapped early-stage founders, they will churn on price and setup complexity no matter how good the software is. The fix lives in your targeting, not your codebase.

The Fix: 3 Tactical Playbooks for Link 1

If your cohort curves do not flatten, pause all top-of-funnel marketing and run these three playbooks.

Tactic 1: Run the “Very Disappointed” Survey

Survey the active users who have been through the product at least twice, and ask one question:

“How would you feel if you could no longer use [product]?”

• A) Very disappointed
• B) Somewhat disappointed
• C) Not disappointed, it isn’t that useful
• D) N/A, I no longer use it

Target metric: 40% or more answering “very disappointed”.

How to use the results: the segmenting matters more than the benchmark. Filter out everyone who answered “somewhat” or “not disappointed”, look only at what the very-disappointed cohort loves and uses most, and make those specific features central to the entire experience.

Tactic 2: Shift from Feature Usage to a Core Habit Metric

Track retention on recurring core actions, not page views or logins.

1. Identify the core action. What action proves the user received value? Slack: messages sent. Figma: files edited. Stripe: transactions processed.
2. Define the natural frequency. Is your product built for daily, weekly, or monthly use?
3. Set the retention threshold. An active user is someone who completes the core action at the natural frequency.

The difference between the two kinds of metric is the difference between comfort and truth:

• [ Vanity metric ] User logged in 3 times this month.
• [ Retention metric ] User exported 2 reports per week for 4 consecutive weeks.

When a user stops hitting the core action, reach out manually within 48 hours. Not an automated email: a direct message from the founder. “Hey, noticed you haven’t run a report this week. Did you hit a bug or a wall?”

Tactic 3: Build Stored Value and Data Moats

The cleanest defensible retention comes from stored value, the data and assets a user accumulates inside your tool that make leaving expensive.

• Historical data: analytics, past logs, search history.
• System integration: connections into the rest of their stack so workflow data flows automatically.
• Customisation and setup: saved templates, custom workflows, team permissions shaped over months.

Then ask the diagnostic question: is your product more valuable to a user after six months than it was on Day 1? If the answer is no, you are asking people to re-decide to stay every single month, from scratch.

Real-World Example: How Retention Saved Amplitude

Early on, the product analytics platform Amplitude had healthy acquisition paired with alarming churn.

Instead of spending more on sales and marketing, they froze growth work and compared cohort data from their small base of sticky users against the users who left.

They found one behavioural difference: users who used team collaboration and dashboard sharing retained at a drastically higher rate than users who built charts alone.

Amplitude rebuilt its product strategy around collaboration, with shared workspaces, team annotations and shared link previews. Retention flattened, which gave them the foundation to scale into a multi-billion-dollar company.

The Takeaway: Earning the Right to Grow

Fixing retention is grueling, because it requires honest conversations with people who stopped using the thing you built. It is also the foundation the rest of distribution stands on.

A product with 5% monthly churn loses nearly half its customer base every year. To grow at all, you have to replace half the business annually just to stay flat. At 2% you lose about a fifth. Same product, same sign-ups, a completely different company.

Fix Link 1 first:

• Measure your cohort retention curves until they flatten.
• Focus relentlessly on the “very disappointed” cohort.
• Build the product around a compounding habit loop.

Once your bucket holds water, you have earned the right to work on Link 2: Messaging, so that the people arriving are the ones most likely to stay.

What to Read Next

• The full Distribution Chain framework: how Retention fits into the complete 7-link system.
• Link 3: Activation: how to get sign-ups to the aha moment once your bucket holds water.