2 builders who never spoke to each other

Read revenue by plan, not in aggregate

An average revenue per user figure hides a team plan billed to one account, and a headline customer count hides the fact that a third of customers can generate two thirds of the money. Break every metric down by tier, then retire or rework the plans contributing almost nothing. Expect user churn to run above revenue churn when your cheapest accounts are the ones leaving, and expect the package mix to decay after a launch as the expensive tiers stop selling.

The method

1. 01

   Work out what share of revenue your cheapest tier actually contributes before you optimise for it.

2. 02

   Expect user churn to exceed revenue churn when your cheapest customers are the ones leaving, and read the gap as a statement about who you are retaining.

3. 03

   Move average revenue per user up by attracting customers who are already bigger, retaining the high end, and adding a pro plan, but not so far that a handful of customers own your business.

4. 04

   Chase net negative churn: the point where existing customers' upgrades outweigh cancellations and downgrades, so revenue would grow even with signups switched off.

5. 05

   Break every metric down by plan, then retire or rework the tiers that contribute almost nothing and consider whether your effort belongs with the tier that dominates.

What it returned

• 20,812

  20,812 customers produced $1,188,580 MRR; the 6,994 paying more than $29 (33% of customers) generated $811,117, or 68% of revenue, with individual customers paying from $29 up to $5,000

• $57

  ConvertKit at $57 per customer against Buffer's $21; his target range was $75-100

• 2.2%

  ConvertKit was at 2.2% net revenue churn with a goal of reaching 0%

• At ConvertKit beginner customers churn more than large ones, so user churn always runs above revenue churn

Sources

• Making $12,500 in 3 months with a book on ember.jsNathan Barry · nathanbarry.com · How were the revenue and sales from the book? · 2015-06-03
• Understanding ConvertKit’s open metricsNathan Barry · nathanbarry.com · The 80/20 of customers · 2019-01-14
• SparkToro's Year One Retrospective - SparkToroRand Fishkin · sparktoro.com · How Did Our First Year Go? · undated
• AI Employees: Named Personas, Cautionary Tale, Outcome Pricing - Trends.vcDru Riley · trends.vc · 🔮 Predictions · 2026-05-21
• Digging in to the Open Startups List - Tyler TringasTyler Tringas · tylertringas.com · MRR contribution by plans, Number of plans and price discrimination · 2015-04-24