---
title: Hard vs soft paywalls
slug: hard-vs-soft-paywalls
summary: When to force users to pay or leave (hard paywall) versus letting them close it (soft), with a case where adding the X halved revenue.
domain: mobile-apps
tags: [monetization, paywall, hard-paywall, product-market-fit]
related: [paywall-in-onboarding, hard-paywall-by-traffic-source, paywall-exit-offers, subscription-conversion-benchmarks, app-growth-stages-organic-to-paid, app-mvp-single-problem-scope, freemium-vs-hard-paywall-free-tier, net-revenue-after-refunds-experiment-metric, removing-the-free-trial]
sources: [yt:vZXn_BSi5AE, yt:2soA7LyyQ64, yt:bg0_DavI-DE, yt:2jH64ziB7uM, yt:L-W6tm-Yk5E, yt:-5zgc14t2gc]
updated: 2026-10-24
---

# Hard vs soft paywalls

## TL;DR
- Test a hard paywall (no X) when traffic is paid, when you get downloads but no sales, or at launch.
- Hard paywalls are common among big health & fitness apps — be aggressive there.
- In one acquired app, a hard paywall + higher price took revenue from $1–2k to $8–10k/month; forcing an X back in halved it.
- The cost: one-star reviews, especially from users who used to get the app free.
- "Hard paywall" means three different things — **no free tier**, **no X**, **no free trial**; RevenueCat's "5x better" stat (10.7% vs 2.1% paid by day 35) only covers the first ([[freemium-vs-hard-paywall-free-tier]]).
- Removing the X can backfire: at BoldVoice trial starts rose but **refunds rose more** — judge on net revenue after refunds ([[net-revenue-after-refunds-experiment-metric]]).
- Middle ground: hard for paid traffic only ([[hard-paywall-by-traffic-source]]) or soft with strong exit offers ([[paywall-exit-offers]]).

## Details
Steve's three cases for a hard paywall:[^1]
1. **Paid traffic** — you're paying for the user, so make them pay or leave.
2. **Downloads but no sales** — ASO/organics work but nobody buys; a hard paywall tests product–market fit immediately because users must start a trial or leave.
3. **App launch** — instead of building feature after feature, launch with a hard paywall to validate demand (especially in health & fitness).

**Case study (Steve's portfolio app):** an acquired app with a soft paywall and a $3 one-time purchase was relaunched with an onboarding flow, a hard paywall and a $60/year subscription — no new features. Revenue went from ~$1–2k to ~$8–10k a month. When the X was added back (Steve says Apple forced it), revenue halved: trial-to-paid stayed the same, but trial-start rate dropped by half — a Mixpanel alert showed trial activation falling from 13% to 6%.[^2][^3][^4]

Downside: the hard version drew one-star reviews from existing users who felt the app used to be free, which is why Steve recommends showing it to paid traffic only.[^2] Hard paywalls guarantee higher trial activation, but watch trial-to-paid, which can fall.[^5]

Benchmark: with a hard paywall Steve expects ≥10% install-to-trial, vs 5–10% for typical apps; he argues for a hard paywall on all Meta traffic.[^6]

**Three meanings (David Barnard, Sub Club).** (1) No free tier — the RevenueCat 2026 report's 10.7% vs 2.1% day-35 conversion compares only this; (2) no X button; (3) no free trial. Each needs a different test: what is a free user worth / can you afford to find out; net revenue after refunds; and whether onboarding builds enough trust to sell without a trial ([[removing-the-free-trial]]).[^7][^8]

**No-X test at BoldVoice.** Removing the X, and separately an X leading to an undismissable discount, both lifted trial starts but refunds spiked by more; BoldVoice now lets users close the paywall into a limited free day one (speech assessment + first practice set), with a trial required from day two.[^9][^10]

## Disagreements / caveats
- Steve's wins were judged on trial starts/revenue; BoldVoice's experience shows refunds can erase a no-X lift — measure on matured net revenue.
- Steve says Apple "forced" the X in his case; App Review enforcement of hard paywalls varies — check current guidelines.
- Single-app anecdote; category matters (fitness tolerates hard walls; utilities with existing free users may revolt).

## See also
- [[paywall-in-onboarding]]
- [[subscription-conversion-benchmarks]]

[^1]: yt:vZXn_BSi5AE @ 01:52 — Steve P. Young
[^2]: yt:vZXn_BSi5AE @ 03:02 — Steve P. Young
[^3]: yt:2soA7LyyQ64 @ 01:58 — Steve P. Young
[^4]: yt:2soA7LyyQ64 @ 02:45 — Steve P. Young
[^5]: yt:bg0_DavI-DE @ 06:17 — Steve P. Young
[^6]: yt:2jH64ziB7uM @ 03:09 — Steve P. Young
[^7]: yt:L-W6tm-Yk5E @ 01:14 — David Barnard
[^8]: yt:L-W6tm-Yk5E @ 08:53 — David Barnard
[^9]: yt:-5zgc14t2gc @ 22:34 — Anada Lakra
[^10]: yt:-5zgc14t2gc @ 24:10 — Anada Lakra