Frictionless SaaS Chapter 13: SaaS Metrics, Cohort Analysis, and the North Star

This is the thirteenth post in the Frictionless SaaS blog series. In Chapter 12 we built the system for catching users before they churn and winning them back. This chapter shifts to measurement: how to know whether any of your retention work is actually moving the needle, and which metrics to obsess over at which stage of your company.

The Metric Problem Almost Every SaaS Team Has

Most SaaS companies track too many metrics and too few meaningful ones at the same time. Dashboards overflow with charts that nobody acts on. Meanwhile the handful of numbers that actually predict whether the business survives sit in a spreadsheet that gets updated quarterly.

Chapter 13 introduces a hierarchy for fixing this — the SaaS Metrics Pyramid — along with the cohort analysis discipline and North Star framing that separate teams who optimize on purpose from teams who optimize on hope.

The core reframe: metrics aren't all equally important at all times. Which metrics deserve your attention depends entirely on which layer of the pyramid is broken. Optimizing the wrong layer is worse than optimizing nothing.

The SaaS Metrics Pyramid

The pyramid has three layers, and the order matters:

Foundation — Activation, Retention, Churn. These are the core health metrics. Without them, nothing else matters. If activation is low, you don't have product-market fit. If churn is high, your business model is broken. The book is explicit about the numbers:

• Healthy B2B SaaS monthly churn sits between 2% and 5%.
• Above 10% monthly churn, the business is not viable regardless of how fast you grow.
• Activation under 30% means your first priority is the onboarding and first-session work from earlier chapters — nothing else.

Middle — Expansion and Growth. Net Revenue Retention (NRR), expansion revenue, upsell rate, cross-sell rate. These metrics measure whether existing customers are growing with you. The book's bar:

• NRR below 100% is a death spiral — churn is outpacing expansion.
• NRR at 100% is treading water.
• NRR above 120% is exceptional and is one of the strongest signals a SaaS business can show.

Apex — The North Star Metric. One metric that captures the core value your product delivers and aligns every team around the same goal. For a project management tool it might be projects completed on time. For a communication tool, messages exchanged across teams. For an analytics tool, insights generated per customer per month.

The test of a good North Star: if you optimize for it and it moves, retention and revenue should move as a consequence. If your North Star can go up while retention goes down, it's a vanity metric. Pick another one.

The book is blunt about sequencing: you do not get to optimize the North Star until the foundation is stable. A company with 30% activation and 12% monthly churn trying to align around a North Star is rearranging deck chairs. Fix the foundation first.

Cohort Analysis: Where Aggregate Metrics Lie

Aggregate metrics hide the most important patterns in your business. A stable-looking churn rate can be masking a newer cohort collapsing while an older cohort holds up. A rising NRR can be hiding the fact that new acquisition is getting worse. Cohort analysis is the only way to see what aggregate numbers conceal.

The chapter covers the two kinds of cohorts that actually matter:

• Time-based cohorts — users grouped by signup month or week. When the January cohort hits 60% three-month retention but the February cohort hits 45%, something changed — in your product, your onboarding, your acquisition channels, or your messaging. Cohort comparison is how you find the "something."
• Characteristic cohorts — users grouped by segment, plan tier, company size, acquisition channel, or role. This is how you discover that enterprise customers have 95% annual retention while SMB customers have 60%, and you tailor your retention strategy accordingly instead of applying a single playbook to wildly different populations.

The book also covers how to use cohort analysis to diagnose the cause of retention changes — whether the delta comes from product changes, acquisition shifts, onboarding breakage, or external factors like industry downturns — so you can actually fix the problem instead of guessing.

Testing Retention Interventions at the Cohort Level

Every retention improvement you ship should be evaluated against the cohort that experienced it versus the cohort that didn't. The chapter makes this non-negotiable because most retention interventions that look good in theory don't move the needle in practice — and without cohort-level testing, you'll never find out.

The uncomfortable truth: a meaningful fraction of the retention initiatives shipped at most SaaS companies have no measurable impact on retention. Teams only avoid discovering this by never running cohort-level comparisons. The book is very direct: if you're not measuring interventions against the cohort that didn't get them, you're not optimizing — you're hoping.

Benchmarks Are Starting Points, Not Targets

The chapter covers benchmark ranges by product category — B2B SaaS, B2C SaaS, enterprise, free products — so you can tell whether your numbers are competitive or whether you're behind the curve. But it also warns against over-indexing on benchmarks.

A company with 8% monthly churn improving from 10% is winning. A company with 3% churn drifting from 2% is losing ground, even though the absolute number still looks better. Trend beats level — which direction you're moving is more important than where you currently sit.

Expansion Metrics Tell the Long-Term Story

Finally, the chapter covers expansion metrics as the leading indicator of whether your product becomes more valuable to customers over time or whether it peaks early and decays. In the best SaaS businesses, customers in their second year spend 50% more than in their first, and customers in their third year spend even more. This isn't just growth — it's evidence that the product is becoming more central to operations, which is the strongest retention signal there is.

When expansion is flat, customers are getting value but not finding new use cases. When expansion declines over tenure, something in the product is failing to keep up with customers' needs — and that's a red flag that no amount of lifecycle messaging can fix.

— Sho Shimoda

Based on Frictionless SaaS: Designing Products Users Discover, Adopt, and Never Leave (2026).