What share of free signups convert to paid by month N, and are newer cohorts converting faster?
№ 023 · Free-to-paid conversion curve
Definition
MetricCumulative share of free signups converting to paid by month N, by cohort
Unit% of cohort signups converted
Percent of free signups converted to paid by month N, one line per cohort. Newer cohorts should sit higher. Build from product analytics joined with billing: count signups per cohort, count converters at each month, divide. Example: Q1 24 cohort hits 16% at M12; Q1 25 cohort already 18% at M6.
Benchmarks
| Bottom 30% | Median | Top 30% |
|---|---|---|
| Opt-in trials under 2.5% (a fifth of trial products sit here); freemium under 2% | Opt-in trials ~8-9%; opt-out trials ~25-31%; freemium ~2.6-5.6% | Opt-in trials 15-25%+; opt-out (card-required) trials 40%+; freemium 8%+ |
B2B SaaS self-serve funnels, 2020-2026 data; conversion measured signup-to-paid, distributions are strongly bimodal so the median is a weak anchor · Model determines the number: requiring a card up front filters intent and multiplies per-signup conversion 3-4x while shrinking signup volume. The 2026 ChartMogul/ProductLed study of 200 products found a 10x spread between top and bottom quintiles and a bimodal shape (20% below 2.5%, 23% above 25%), so compare against your model and cluster, not the blended median. Denominator hygiene matters: all-signups vs activated-users bases differ by multiples.
| Category | Bottom 30% | Median | Top 30% |
|---|---|---|---|
| By conversion model | opt-in under 2.5%; freemium under 2% | opt-in ~8-9%; opt-out 25-31%; freemium 2.6-5.6% | opt-in trial 15%+; opt-out trial 40%+; freemium 8%+ |
When it looks bad
Newer cohort curves start lower and never catch up: each vintage's cumulative line sits below the previous one at every month N, meaning signup quality or first-run experience degraded and the funnel is filling with users who were never going to pay.
Jan cohort reached 9% by month 3; April cohort 6.5%; July cohort tracking to 4.8%. Signup volume grew 40% over the same window, so the dashboard's absolute conversions look flat while the engine decays.
What to do about it
- Split the curve by acquisition source before touching the product: a falling blended curve with stable organic conversion means paid or partnership traffic quality dropped; fix targeting, not onboarding.
- Instrument one activation milestone and measure activated-to-paid separately: if headline conversion is 3% but activated users convert at 12%, the fix is activation flow, not pricing (Artisan framework); gate lifecycle emails on the milestone, not on days elapsed.
- Convert at the moment of limit: usage-limited free tiers with an in-app upgrade prompt at the limit convert measurably better than feature-gated tiers; pair the prompt with a first-month offer.
- Choose the model for the economics, not the benchmark: card-required trials triple per-signup conversion but cut volume; run the 1,000-visitors math (OpenView) on your own traffic before switching models.
Sources
- Userpilot (reporting the Kyle Poyar x ChartMogul x ProductLed study) Free-trial median ~8%; bimodal: 20% of products below 2.5%, 23% above 25%; 10x gap between top and bottom self-serve quintiles userpilot.com ↗
- Pulseahead (reporting ChartMogul 2026 study data) Opt-in trials 8.9% average; opt-out (card required) 31.4%; freemium 5.6%; +1 point of conversion = ~15% more new revenue per cohort pulseahead.com ↗
- OpenView Median trial conversion ~2x freemium (roughly 14% vs 7%); freemium converts more of total site traffic despite lower per-signup rates openviewpartners.com ↗
- Artisan Growth Strategies (aggregating OpenView, ProfitWell, ChartMogul, 1Capture data) Freemium self-serve good at 2-5%, sales-assisted 5-7%, above 10% excellent; median B2B trial-to-paid 18.5% in the aggregated dataset artisangrowthstrategies.com ↗
- Product Growth Blog (reporting OpenView/High Alpha and Growth Unhinged data) B2B freemium median ~2.6%, top quartile 5-8%; AI-native products run hotter (good 6-8%, great 15-20%) productgrowth.blog ↗