---
title: Paywall visuals — video, photos and social proof
slug: paywall-visuals-and-social-proof
summary: Video beats photos beats graphics on paywalls (one app went 2.5% → 4.7% → 6.8%), social proof helps, and a money-back guarantee line was winning in a RevenueCat experiment.
domain: mobile-apps
tags: [monetization, paywall, design, social-proof, video]
related: [long-paywall-sell-the-outcome, three-screen-pre-paywall-sequence, revenuecat-charts-and-experiments, better-sleep-onboarding-paywall-teardown, paywall-design-learnings-at-scale]
sources: [yt:vZXn_BSi5AE, yt:2soA7LyyQ64, yt:hs_hZS__hvA, yt:2oq2OlqfTro, yt:FWF_AbsR0qc, yt:8afSYJU-y8s, yt:lkwX_kc0NS8]
updated: 2026-10-24
---

# Paywall visuals: video, photos, social proof

## TL;DR
- Hierarchy Steve has seen repeatedly: **video > photos of real people > graphics/illustrations**.
- Momingo (transit app, soft paywall): graphics 2.5% → real person 4.7% → feature video 6.8% conversion.
- Add social proof on or right before the paywall (ratings, user counts, testimonials) — it beat no social proof in Steve's tests.
- Test a money-back-guarantee line: it was beating control on initial conversions in a live RevenueCat experiment.
- A "preparing your plan" screen with animation + social proof before the paywall: +10% reported by one founder.
- Superwall (Jonathan Parra): **comparison tables underperform** bullets/video (a simpler image-led variant won +111%); **Rotato** phone-mockup videos + social proof + bullets win broadly; a **native SwiftUI look** beat custom designs in two apps; a **dark, premium style** for the top tier nudged users to Pro.

## Details
**Media.** John (Momingo founder) shared on a livestream: graphics paywall converted 2.5%, swapping in a real person 4.7%, adding a video showing the features 6.8% — with a soft paywall.[^1] Steve's summary rule: photos better than graphics, video better than photos.[^2] Momingo also found photos worked better on the paywall itself.[^3]

A client copied Cal AI's flow (video paywall plus extra screens behind "continue"); when tested, the **simple video paywall won** over the longer variant.[^1][^4]

**Social proof.** In the two-paywall test, the trial paywall had social proof at the bottom; Steve has seen this outperform versions without it.[^5] Momingo's onboarding repeats proof on every screen ("over 5 million riders", "50,000 five-star" ratings, quotes).[^3] Better Sleep shows a plan-preparation screen with "trusted by 65M people, 91% sleep better, 4.8 stars, featured 65 times" while it "builds your plan"; another founder told Steve that adding animation before the paywall raised conversion 10%.[^6]

**Guarantee.** Steve's team ran a RevenueCat Experiment adding money-back-guarantee language to a custom paywall; after four days variant B led on initial conversions and trial starts (not yet significant; he checks LTV/revenue at the bottom of the report before calling it).[^7]

**Superwall design data.** Parra: tables lose to simple bullets, video or a product visual because users don't read; one table paywall lost to an image-led "We want you to try [app] for 50% off" variant by 111%.[^8] His video paywalls use Rotato to put a screen recording into a phone mockup, with social proof and bullets; a Mojo-inspired one won its test.[^9][^10] In apps built in stock SwiftUI, a SwiftUI-styled paywall outperformed custom ones "by quite a long shot".[^11] For two-tier apps, switching the paywall to a dark high-contrast style when Pro is selected (Hinge-inspired) increased Pro uptake.[^12] Social proof works especially well for large apps with many reviews.[^13]

## Disagreements / caveats
- Sample sizes are small and mostly anecdotal; the guarantee test was only four days old at recording.
- Video costs load time; test on slow networks.

## See also
- [[long-paywall-sell-the-outcome]]
- [[revenuecat-charts-and-experiments]]

[^1]: yt:vZXn_BSi5AE @ 10:00 — Steve P. Young
[^2]: yt:2soA7LyyQ64 @ 18:00 — Steve P. Young
[^3]: yt:FWF_AbsR0qc @ 08:31 — Steve P. Young
[^4]: yt:vZXn_BSi5AE @ 10:55 — Steve P. Young
[^5]: yt:hs_hZS__hvA @ 01:52 — Steve P. Young
[^6]: yt:8afSYJU-y8s @ 05:05 — Steve P. Young
[^7]: yt:2oq2OlqfTro @ 07:43 — Steve P. Young
[^8]: yt:lkwX_kc0NS8 @ 25:40 — Jonathan Parra
[^9]: yt:lkwX_kc0NS8 @ 44:01 — Jonathan Parra
[^10]: yt:lkwX_kc0NS8 @ 45:11 — Jonathan Parra
[^11]: yt:lkwX_kc0NS8 @ 47:15 — Jonathan Parra
[^12]: yt:lkwX_kc0NS8 @ 52:04 — Jonathan Parra
[^13]: yt:lkwX_kc0NS8 @ 34:18 — Jonathan Parra