+150 XP

Cohort economics and net revenue retention as the growth engine

# Cohort economics and net revenue retention as the growth engine

Picture two SaaS companies that both win 100 new customers a month. Company A grows revenue 40% a year. Company B is flat. Same sales team, same product category, same price point. The difference is not on the acquisition side at all. It is what happens to the customers after they sign, month after month, in the cohorts nobody looks at once the deal closes.

That hidden layer is where SaaS finance actually lives.

The cohort: your unit of truth

A cohort is a group of customers who started in the same period, usually the same month. The January 2026 cohort is everyone who first paid you in January 2026.

Why group them this way? Because blended, company-wide revenue hides everything. If total revenue is up, you cannot tell whether existing customers are expanding, or whether you are simply pouring new sales on top of a leaky bucket. Cohorts separate the two.

Track each cohort's revenue over time and you get a retention curve: how much of that starting group's revenue you still have in month 3, month 12, month 24.

Two things pull that curve in opposite directions:

  • Gross churn: revenue lost when customers cancel or downgrade. This only goes down.
  • Expansion: revenue gained when existing customers upgrade, add seats, or buy more usage. This can push a cohort *above* where it started.

Net revenue retention, defined

Net revenue retention (NRR) measures how much revenue a cohort generates now versus 12 months ago, counting only existing customers. New logos are excluded on purpose.

The formula, over a 12-month window:

NRR = (Starting MRR + Expansion - Contraction - Churn) / Starting MRR

where:
  Starting MRR  = monthly recurring revenue from the cohort 12 months ago
  Expansion     = upgrades, added seats, usage increases
  Contraction   = downgrades (customer stays, pays less)
  Churn         = revenue from customers who fully left

MRR means monthly recurring revenue, the predictable subscription revenue that renews each month.

Read the output like this:

  • NRR = 100%: existing customers regenerate exactly what they were worth. Expansion offsets churn.
  • NRR below 100%: your existing base shrinks. You are running down an escalator.
  • NRR above 100%: your existing base grows on its own, with zero new customers.

That last case is the whole game.

The 120% cohort that grows itself

Say the January cohort starts at $100,000 in MRR. Twelve months later, with 120% NRR, that same group (minus anyone who left, plus everyone who expanded) is now worth $120,000 in MRR.

You added no new customers to that cohort. It grew 20% anyway.

Now stack cohorts. Layer January on February on March, each one compounding at 120%, and the base revenue climbs even if new-logo sales slow to a trickle. This is the retention waterfall: each period's revenue is last period's revenue, minus churn, plus expansion, before you add a single new sale on top.

This is why investors treat NRR as the single most predictive SaaS metric. A company at 120%+ NRR has a growth engine embedded in customers it already won. Bessemer's widely cited State of the Cloud research has long flagged best-in-class NRR in the 120% to 130% range for top infrastructure and usage-based businesses, though most healthy SaaS companies sit closer to 100% to 110%.

🎬 [VIDEO: "SaaS Metrics Masterclass: Net Revenue Retention" - youtube.com - a clear walkthrough of NRR, gross retention, and the cohort waterfall for finance and operating teams]

Gross churn: the ceiling nobody prices in

Here is the trap. Expansion can flatter your NRR while a serious problem hides underneath.

Gross revenue retention (GRR) strips out expansion entirely. It counts only what you kept: starting revenue minus churn and contraction, with no upside allowed. GRR can never exceed 100%.

GRR = (Starting MRR - Contraction - Churn) / Starting MRR

A company can post 120% NRR while losing 25% of its customers a year. A handful of whales expanding hard masks a base that is bleeding out underneath. NRR looks like a rocket; GRR reveals a sieve.

Why does this cap your acquisition spend? Think of churn as a tax on every dollar you acquire.

If gross churn is 10% a year, you keep 90 cents of every revenue dollar you land. If churn is 30%, you keep 70 cents, and you must re-acquire the lost 30% just to stand still. High churn means your sales and marketing engine spends more of its output replacing lost revenue instead of adding new revenue.

At some churn rate, new sales and lost revenue cancel out. That is your ceiling. No amount of extra ad spend breaks through it, because the bucket empties as fast as you fill it.

A quick numeric feel

Start at $10M in annual recurring revenue (ARR). You add $3M in new ARR this year.

  • 10% gross churn: you lose $1M, net add is $2M, you end at $12M.
  • 30% gross churn: you lose $3M, net add is $0, you end flat at $10M.

Same $3M of hard-won new business. Completely different outcome. The lever was retention, not acquisition.

This is why experienced SaaS CFOs watch GRR before NRR. GRR tells you if the product delivers durable value. NRR tells you how much upside sits on top of that foundation. A high NRR built on a weak GRR is fragile: lose one or two expanding whales and the whole number collapses.

Where the lens matters: usage vs. seat pricing

The pricing model shapes the retention curve.

Seat-based pricing (charge per user) makes expansion depend on customer headcount growth. Expansion is steady but bounded. Contraction hits fast in a downturn when customers cut seats.

Usage-based pricing (charge per API call, per gigabyte, per transaction) ties your revenue to the customer's own growth. When they scale, you scale automatically, no upsell conversation required. This is why usage-based leaders often post the highest NRR numbers: expansion is baked into the meter. The tradeoff is volatility. Usage can drop as fast as it rose.

For finance teams, this changes forecasting. Seat revenue is more predictable month to month. Usage revenue has higher expected NRR but wider variance, which matters for cash planning and for how you talk to investors about revenue quality.

Knowledge check

1. Two SaaS companies acquire the same number of new customers each month, yet one grows 40% annually while the other stays flat. What does this scenario primarily illustrate?

2. Why does analyzing revenue by cohort reveal more than looking at blended company-wide revenue?

3. A cohort's retention curve rises above 100% of its starting revenue over time. What must be true?

MULTIPLE CHOICE

4. Select ALL correct answers about how net revenue retention (NRR) is defined and calculated.

Select all the correct answers.

MULTIPLE CHOICE

5. Select ALL correct answers that correctly distinguish the forces acting on a cohort's revenue over time.

Select all the correct answers.

Reading a cohort table like an operator

When you get a cohort retention table, do not just glance at the headline NRR. Walk it:

1. Find the shape of the curve. Does revenue dip in the first few months (early churn from bad-fit customers) then flatten or rise? A curve that flattens above 100% is a compounding machine. A curve that keeps sliding is a churn problem no marketing budget fixes.

2. Separate GRR from NRR in your head. If NRR is 118% but GRR is 82%, ask who is expanding. Concentration risk lives here. If three accounts drive all the expansion, your growth engine is really three phone numbers.

3. Compare cohorts over time. Is your most recent cohort retaining better or worse than last year's? Improving retention across newer cohorts is one of the strongest signals that product and onboarding are working. Degrading retention is an early warning that shows up long before it hits the headline growth rate.

4. Segment. Enterprise cohorts usually retain far better than small-business cohorts. A blended NRR of 105% might be 130% enterprise and 85% SMB. Those are two different companies wearing one number.

For a solid free primer on the underlying math, the OpenView SaaS metrics benchmarks archive and Christoph Janz's writing on SaaS napkin math are both worth bookmarking.

Why this is the growth engine

Return to Company A and Company B from the opening. Both landed 100 customers a month. Company A retains and expands its base at 120% NRR. Its old cohorts grow while new ones stack on top. Company B sits at 90% NRR: every new cohort partly refills what older cohorts lost.

Company A compounds. Company B runs to stand still.

The acquisition engine gets the attention and the budget. But cohort economics decide whether that spend compounds into durable growth or drains into a leaking base. Retention is not a customer-success side project. It is the core financial lever of the model.

Key Takeaways

  • Cohorts are the unit of truth. Blended revenue hides whether your base is growing or leaking. Group customers by start month and track their revenue over time.
  • NRR above 100% means your existing customers grow on their own. Best-in-class is often cited around 120% to 130%; most healthy SaaS sits near 100% to 110%.
  • Always read GRR alongside NRR. A high NRR built on a weak GRR is fragile and usually hides customer concentration or a churn problem.
  • Gross churn caps your acquisition ceiling. Above a certain churn rate, new sales only replace lost revenue, and more marketing spend cannot break through.
  • Pricing model shapes the curve. Usage-based pricing tends to lift NRR but adds volatility; seat-based pricing is steadier but bounded.