Free tier or not? Freemium vs charging from day one

TL;DR

• RevenueCat 2026 data: apps with no free tier convert 10.7% of users to paid by day 35 vs 2.1% for freemium (~5x); year-one retention of annual subs is a wash (27% vs 28%).
• Decide by asking what a free user is worth to you. If a free user is just a compute/data bill or someone you aren’t building for, go hard paywall.
• Charging from day one doubles as ICP discovery: willingness to pay filters out feedback from people who will never be customers (BoldVoice, $10M ARR, 10 people).
• Freemium pays when free users are the growth engine — word of mouth (Opal) or network effects (Tinder: 85–90% free).
• Switching to freemium is risky: one Phil Carter client lost >50% of subscriber conversion and reverted within weeks. Ask “can I afford to find out?”
• Don’t mix playbooks: copying Duolingo’s freemium base into a high-intent niche app is “the worst thing” (Anada Lakra).

Details

The data. In RevenueCat’s 2026 State of Subscription Apps report, hard-paywall (no free tier) apps converted 10.7% of users to paid by day 35, freemium 2.1%. David Barnard expected forced day-one decisions to raise churn, but year-one retention on annual subscribers was 27% (hard) vs 28% (freemium). Jacob Eiting’s read: users are never as primed as right after search → download → onboarding, so capture it.¹² Note the 5x only compares free tier vs no free tier — not “no X button” or “no trial” (Hard vs soft paywalls, Removing the free trial (pay-up-front plans)).²

When a free user costs money. David’s Weather Up was freemium with banner ads, but every open cost him weather-data fees, ads barely covered it, and users complained about ads. He pulled ads and shipped a hard paywall (grandfathering existing free users, which he’s been losing money on for years).³

Charging to find your ICP. BoldVoice (AI accent coaching) has had a hard paywall since launch at YC in 2021: a simple ~$10/month with a 7-day trial, so the team could focus on product, not pricing experiments. Anada Lakra’s reasoning: paying users reveal who the problem truly resonates with, so you stop getting distracted by feedback from non-ICP users.⁴⁵ BoldVoice deliberately ignores beginner learners (Duolingo’s base) and targets intermediate–advanced career professionals whose alternative is a $400–500/hour coach.⁶⁷ Early, the “obvious” college-student playbook failed: students improve on their own, can’t pay much, and haven’t yet felt the career cost of their accent.⁸

When free users are the marketing. Opal (screen time) reached ~$10M ARR on a hard paywall, then went freemium: paid penetration of MAU fell from ~20% to 9%, but DAUs “exploded” and revenue grew with them. Kenneth Schlenker’s rule for where to draw the line: if a free user wouldn’t recommend your app, give away more.⁹¹⁰

When free users are the network. Dating needed free-to-play to win (OkCupid, Plenty of Fish, then Tinder): about 85–90% of Tinder users are entirely free, because in-demand users (mostly women) must not pay to be there; the 10–15% payers, predominantly men, fund it.¹¹¹² Ads exist but are a small share — Ravi Mehta and Jacob put ads at roughly 10–20% of revenue at most for apps they know.¹³

Switching is hard. Phil Carter (Elemental Growth) calls moving from hard paywall to freemium “checkers to chess”. One client got +75% LTV per user from a multi-step paywall plus pricing/packaging changes; another’s freemium shift cut subscriber conversion >50% and was pulled after a couple of weeks. His view: for most apps — especially bootstrapped or capital-light — hard paywall is the more reliable, lower-risk path.¹⁴

Disagreements / caveats

• The 10.7% vs 2.1% gap is correlational across apps (category and intent differ), not a controlled switch.
• Opal’s freemium win depended on revenue still growing underneath the conversion drop; most small apps couldn’t survive that dip.
• Steve P. Young’s advice (Hard vs soft paywalls) leans hard-paywall too, but on paid traffic; Sub Club sources add that free users can have real value.
• Middle grounds: a limited free day-one experience or reverse trial (Reverse trials — converting users who closed the paywall), metered free usage (What to put behind the paywall (drawing the free/paid line)).

See also

• What to put behind the paywall (drawing the free/paid line) — if you do keep a free tier
• Setting your first subscription price — charge high enough to learn
• David Barnard, Anada Lakra, Ravi Mehta

Footnotes

1. The case for (and against) hard paywalls @ 00:00 — David Barnard, Jacob Eiting ↩

2. The case for (and against) hard paywalls @ 01:14 — David Barnard ↩ ↩²

3. The case for (and against) hard paywalls @ 02:12 — David Barnard ↩

4. Why BoldVoice charged from day 1 and ignored the Duolingo freemium playbook — Anada Lakra @ 13:03 — Anada Lakra ↩

5. Why BoldVoice charged from day 1 and ignored the Duolingo freemium playbook — Anada Lakra @ 14:35 — Anada Lakra ↩

6. Why BoldVoice charged from day 1 and ignored the Duolingo freemium playbook — Anada Lakra @ 29:01 — Anada Lakra ↩

7. Why BoldVoice charged from day 1 and ignored the Duolingo freemium playbook — Anada Lakra @ 29:46 — Anada Lakra ↩

8. Why BoldVoice charged from day 1 and ignored the Duolingo freemium playbook — Anada Lakra @ 33:53 — Anada Lakra ↩

9. The case for (and against) hard paywalls @ 03:30 — Kenneth Schlenker ↩

10. The case for (and against) hard paywalls @ 04:29 — David Barnard ↩

11. How Tinder Captures More Value With Tiered Pricing and Consumables — Ravi Mehta @ 06:39 — Ravi Mehta ↩

12. How Tinder Captures More Value With Tiered Pricing and Consumables — Ravi Mehta @ 08:40 — Ravi Mehta ↩

13. How Tinder Captures More Value With Tiered Pricing and Consumables — Ravi Mehta @ 12:05 — Ravi Mehta, Jacob Eiting ↩

14. The case for (and against) hard paywalls @ 05:23 — Phil Carter ↩