---
title: Subscription funnel benchmarks (trial start, trial-to-paid, install-to-paid, LTV per install)
slug: subscription-conversion-benchmarks
summary: Target numbers for subscription apps — 5–10% install-to-trial (10%+ with a hard paywall), 30–40% trial-to-paid, ~2–3% install-to-paid, and ≥$3 realised LTV per install before scaling Meta.
domain: mobile-apps
tags: [monetization, metrics, benchmarks, ltv, paid-ua]
related: [revenuecat-charts-and-experiments, hard-vs-soft-paywalls, paid-intro-offers, trial-to-discount-offer, meta-ads-readiness-checklist, freemium-vs-hard-paywall-free-tier, removing-the-free-trial, membership-churn-benchmarks-and-ltv]
sources: [yt:2oq2OlqfTro, yt:2jH64ziB7uM, yt:bg0_DavI-DE, yt:2soA7LyyQ64, yt:24QMRDXck70, yt:lkwX_kc0NS8, yt:L-W6tm-Yk5E, yt:gqN6Z5WAeiI]
updated: 2026-10-24
---

# Subscription funnel benchmarks

## TL;DR
| Metric | Okay | Good | Great |
|---|---|---|---|
| Install → trial start ("start rate") | 5–10% | >10% (expected with hard paywall) | >20% (rare) |
| Trial → paid | ~30–40% (RevenueCat iOS median ~33%) | >40% | 60%+ (niche 7-day yearly) |
| Install → paid | ~2% | 2.6% (RevenueCat iOS median by day 35) | — |
| Realised LTV per install (for Meta) | ≥ $3 | higher | — |
| $1.99 paid intro → renewal | 60–70% (Steve) | | |

- Vahe Bagdasarian (Superwall agency) sets the bar higher: **trial start ≥15%** is good (8–11% is "easily" 15%), trial-to-paid ≥30%; **no-trial apps (US traffic): install-to-paid ≥10% good, 4–10% okay, <4% problem**.
- RevenueCat 2026: no-free-tier apps convert **10.7%** of users to paid by day 35 vs **2.1%** for freemium.
- Niche apps get higher start and trial-to-paid rates.
- First questions Steve asks a new client: what's your trial start rate and your trial-to-paid?

## Details
**Start rate & trial-to-paid.** In RevenueCat's Trial Conversion chart (set to *relative*), Steve's niche health & fitness app shows a high start rate (aiming for 12%; January peaks for fitness) and >60% trial-to-paid on a 7-day trial. His guide: 5–10% start rate is good, >10% great, >20% amazing and rare; trial-to-paid averages 30–40%. The more niche the app, the higher both numbers.[^1][^2]

**Before Meta ads** Steve wants: ≥5–10% install-to-trial (≥10% with hard paywall), >30% trial-to-paid, ~2% install-to-paid. Worked example: 1,000 installs × 7% = 70 trials × 30% = 21 subscribers ≈ 2.1% paid.[^3][^4] RevenueCat's State of Subscription Apps: 2.6% median iOS download-to-paid by day 35, ~33% median iOS trial-to-paid — compare within your category and country.[^5]

**LTV per install.** Steve (and his friend Henry, who'd want more) requires at least ~$3 realised LTV per install to run Meta; check the curve at month 1 vs month 11 — for some apps it barely moves.[^6][^7] In RevenueCat, Steve's app sits at ~$3 per customer at 30 days and barely more (~$3.3) at 12–24 months.[^8] Levers: conversion rates and pricing.[^9]

**Retention context:** the average app loses ~77% of DAUs within 3 days; day-30 retention ~4% (AppsFlyer).[^10]

**Free-user conversion:** typical free → paid ≈ 2%; Momingo's reverse trial got 4.5%.[^11][^12] Paid intro offers convert/renew at ~60–70% vs 30–40% for free trials.[^13]

**Superwall-side benchmarks.** Vahe: trial start above 15% is good, below 15% means work to do; trial-to-paid above 30%; for no-trial apps with mostly US users, ≥10% install-to-paid is very good, 4–10% "not terrible, not great", <4% a problem.[^14][^15] Jonathan Parra sees ~8% paywall conversion as the cross-app average, reaching 15–20% after a few iterations.[^16] Examples: Claim reached 34% trial start and 50–55% trial-to-paid; an entertainment app went 1.1% → ~4% install-to-paid in five months on paywall work alone.[^17][^18]

**Free tier vs none.** RevenueCat's 2026 report: hard-paywall apps convert 10.7% of users to paid by day 35, freemium 2.1%; year-one retention of annual subs 27% vs 28%.[^19][^21]

**High-trust outlier.** Natal (founder-led fitness, $25/month) reports ~17% download-to-trial vs ~6% health & fitness average and ~68% trial conversion vs a ~38% average she cites.[^20]

## Disagreements / caveats
- Vahe's "15% trial start" bar is much higher than Steve's 5–10% — Vahe's clients are often hard-paywall, multi-step-paywall apps; definitions (per install vs per paywall view) may differ.
- Steve's figures are rules of thumb from his clients; RevenueCat medians are the more defensible public reference (2026 report as cited).
- Trial-to-paid alone misleads with hard paywalls (start rate rises, trial-to-paid may fall).

## See also
- [[revenuecat-charts-and-experiments]] — where to read these numbers

[^1]: yt:2oq2OlqfTro @ 04:01 — Steve P. Young
[^2]: yt:2oq2OlqfTro @ 04:49 — Steve P. Young
[^3]: yt:2jH64ziB7uM @ 03:09 — Steve P. Young
[^4]: yt:2jH64ziB7uM @ 04:02 — Steve P. Young
[^5]: yt:2jH64ziB7uM @ 04:54 — Steve P. Young
[^6]: yt:2jH64ziB7uM @ 06:37 — Steve P. Young
[^7]: yt:2jH64ziB7uM @ 07:35 — Steve P. Young
[^8]: yt:2oq2OlqfTro @ 05:38 — Steve P. Young
[^9]: yt:2oq2OlqfTro @ 06:31 — Steve P. Young
[^10]: yt:bg0_DavI-DE @ 06:17 — Steve P. Young
[^11]: yt:bg0_DavI-DE @ 04:36 — Steve P. Young
[^12]: yt:2soA7LyyQ64 @ 08:08 — Steve P. Young
[^13]: yt:bg0_DavI-DE @ 08:02 — Steve P. Young
[^14]: yt:24QMRDXck70 @ 04:32 — Vahe Bagdasarian
[^15]: yt:24QMRDXck70 @ 05:26 — Vahe Bagdasarian
[^16]: yt:lkwX_kc0NS8 @ 55:02 — Jonathan Parra
[^17]: yt:24QMRDXck70 @ 16:28 — Vahe Bagdasarian
[^18]: yt:24QMRDXck70 @ 06:16 — Vahe Bagdasarian
[^19]: yt:L-W6tm-Yk5E @ 00:00 — David Barnard
[^21]: yt:L-W6tm-Yk5E @ 01:14 — David Barnard
[^20]: yt:gqN6Z5WAeiI @ 25:39 — Nancy Anderson