B2B SaaS retention benchmarks 2026 (by stage, cohort, and segment)

Most founders compare their retention to nothing.

They look at the aggregate number, decide it's "okay", and move on. The dashboards confidently show 30% twelve-month retention. The founder feels neutral about it. Three quarters later, the cohorts are visibly bleeding.

Without benchmarks, the number is meaningless. With benchmarks, the number is a diagnostic.

This is the cluster post for foundation one of the pillar: The 7x rule: why retention beats acquisition. If you're guessing whether your retention is good, this is the diagnostic.

Why aggregate retention is the most-lied-about number

Aggregate retention is the average. Averages hide everything that matters.

A founder with 30% aggregate twelve-month retention could have:

• A March 2025 cohort with 40% retention, an October 2025 cohort with 22% retention, average 31%. The trend is collapsing.
• Or every cohort at 30-32% retention, very stable. The business is healthy.

Same aggregate number. Opposite stories. The aggregate hides which.

The fix is cohort visibility. Every founder at €30K-€500K MRR should be able to pull, in under 5 minutes:

• 12 most recent monthly cohorts
• Retention at 1 / 3 / 6 / 12 months post-signup for each
• The trend direction across cohorts

If you can't pull this, the next 30 minutes are your project.

The four cohort views every founder should pull

1. Month-over-month retention by signup cohort. Rows are signup month. Columns are months since signup. Cells are percentage active.
2. Revenue retention by cohort. Same shape, but cells are revenue retained (accounts for upgrades and downgrades).
3. Segment-cut retention. Split your customer base by segment (free vs paid, SMB vs mid-market, channel of acquisition). Rerun the cohort table per segment.
4. Activation-conditional retention. Filter cohorts to only users who hit your activation event. Compare to non-activated. This isolates retention from acquisition quality.

If you don't have a defined activation event, that's a prerequisite — see the activation event 7-day definition.

Benchmark ranges — B2B SaaS

Bracket your retention against these (12-month, post-activation):

• Best-in-class (top 10%): 80%+ logo retention, 100%+ net revenue retention (NRR)
• Healthy: 70-80% logo retention, 95-105% NRR
• Median: 60-70% logo retention, 85-95% NRR
• Below median: 50-60% logo retention, 75-85% NRR
• Structural problem: under 50% logo retention, under 75% NRR

NRR > 100% (revenue grows from existing customers even after churn) is the SaaS gold standard. Achievable when upsell + expansion outpaces churn + downgrade.

Benchmark ranges — Ecom / DTC

The math is different because of repeat-purchase economics.

• Best-in-class: 35%+ year-2 repeat customer rate, 4+ orders per repeat customer
• Healthy: 25-35% year-2 repeat rate
• Median: 15-25%
• Below median: under 15%

For ecom, "retention" is "did they buy again". The math:

• CLV: average customer lifetime value
• CAC payback: months to break even on acquisition cost
• Order frequency: orders per year per customer

Healthy ecom at €30K-€500K MRR: 25%+ year-2 repeat rate AND CAC payback under 6 months.

Benchmark ranges — Subscription (non-SaaS)

Subscription boxes, content subs, fitness, etc:

• Best-in-class: 70%+ month-3 retention, 50%+ month-12 retention
• Healthy: 60-70% month-3, 35-50% month-12
• Median: 50-60% month-3, 25-35% month-12

The first three months are where most subscription businesses leak. Month 4+ tends to flatten.

How to compare yourself honestly

Three rules:

1. Compare to your segment, not the industry average. "SaaS retention" hides massive variance between SMB SaaS (lower retention) and enterprise SaaS (higher).
2. Use cohort, not aggregate. Aggregate misleads. Cohort exposes.
3. Track trajectory, not just snapshot. Improving cohort retention quarter-over-quarter matters more than the absolute number.

If your most recent cohort is significantly worse than your cohort from 12 months ago, the retention engine is decaying. That's an emergency. The aggregate number won't tell you for another 12-18 months.

What "good" looks like by MRR stage

At each MRR stage, the retention math changes:

• €10K-€30K MRR (early traction). Retention is mostly product fit. Aim for "no obvious decay" quarter-over-quarter. Don't optimize aggressively yet.
• €30K-€100K MRR (foundation phase). Install cohort visibility, define activation, install the 4 lifecycle flows. Aim for retention to STABILIZE.
• €100K-€500K MRR (scaling). Retention should be a tracked OKR. Quarterly experiments to push it 2-5 points up. NRR > 100% is the target if SaaS.
• €500K+ MRR. Retention is the primary growth lever. More resources here than acquisition.

If you're at €30K-€100K MRR with cohort retention you've never actually pulled, your week-one project is clear.

Where to start

Open your analytics today. Build the monthly cohort table for the last 12 cohorts. The exercise takes 90 minutes and rewires what you see when you look at growth.

If your retention is below median and you don't know why, take the audit. Diagnosis is the first 30 days. Five minutes, auto-qualifies fit.

→ Take the 5-min audit

What's your most recent cohort's 3-month retention, and how does it compare to your 12-month-ago cohort?