Guides
6
Churn rate: three versions, and which one to watch
Customer churn, revenue churn and net churn - how each is calculated, why they diverge, and what counts as normal at small scale.
3 min read
How to calculate MRR, and four ways it goes wrong
How to calculate monthly recurring revenue correctly, and the four accounting mistakes that make the number look better than it is.
2 min read
Calculating marketing ROI when the payback takes months
How to calculate marketing ROI when payback takes months, which costs belong in the denominator, and how to handle attribution honestly.
3 min read
The LTV:CAC ratio, and why 3:1 is not always the goal
How to calculate LTV:CAC without fooling yourself, what the ratio hides, and why CAC payback period is the more useful number for a bootstrapped product.
3 min read
SaaS metrics explained: MRR, ARR, churn, LTV and CAC
How each core SaaS metric is calculated, the definitional traps that make them wrong, and which ones actually matter before $10k MRR.
3 min read
The SaaS pricing page: what to show, what hiding costs
How many tiers, where to anchor, whether to show prices at all, and the specific patterns that lose signups on a pricing page.
3 min read
About saas metrics
Every SaaS metric has at least two definitions in common use, and the gap between them is usually larger than the change you were trying to measure. Churn can be counted by customer or by revenue, monthly or annually, with or without downgrades, and the same business can honestly report three very different numbers. This hub is about picking one definition, saying which one, and sticking to it.
The guides work through MRR, ARR, churn, LTV, CAC and payback period the same way each time: the formula, the inputs people get wrong, what the number is useful for, and at what stage it starts being meaningful. Several of them are not meaningful early. Lifetime value computed from four months of data is a projection with an enormous error bar, and treating it as a fact leads to spending decisions that do not survive the year.
Benchmarks appear where honest ones exist, with the caveat that published benchmarks skew heavily toward companies that had a reason to publish. A number that looks bad against a public benchmark is often fine for your segment and stage.
Common questions
- Which churn number should be reported?
- Report both logo churn and revenue churn, monthly, and say which is which. Revenue churn can be near zero or negative while logo churn is high if expansion offsets losses, and one number alone hides that.
- When does LTV become useful?
- Once you have enough history that observed retention, not an assumed curve, drives the calculation. Before that it mostly restates your assumptions back to you.