Growing to $1M MRR with Paywall and Pricing Experiments — Francescu Santoni, Mojo

After the pivot, took about two months (with two founders plus an intern) to build a glitchy alpha and launched it on Product Hunt. Added a subscription from day zero as the demand signal: $10/month (they had considered $1) with a 30-day free trial so nobody paid by accident. No onboarding paywall for the first few years, so trials reflected actual interest. Success bar set in advance: one trial from a country other than France.

After Apple's App Store team told them they were not pushing the paywall enough, the first paywall experiment moved the existing paywall, unchanged, to the end of onboarding. The team had resisted because users had not tried the app yet, but users get a free trial first.

One dedicated monetization hire (Michal, data-minded, documents everything) owns paywall, pricing and onboarding experiments using Superwall and RevenueCat, borrowing a developer only when needed. One focus per quarter, in order: (1) paywall position, (2) totally different paywall concepts (video, minimal text, scrolling, trial reminder/timeline), (3) localized paywalls per country, (4) smaller changes within the winning paywall, (5) price tests.

Benchmarked prices of video and photo editing apps and chose to charge more than most, targeting prosumers such as small businesses and agencies (some users run agencies on Mojo with freelancers), since casual users give weak signal in a freemium app. Raised prices through tests; annual price now around $60-70. Accepts complaints in reviews and interviews from casual users.

Noticed Turkish inflation had eroded the local price point. The monetization lead tested a +100% price in Turkey and then rolled it out.

Mojo invented animated posters in Instagram stories, a format that did not exist. When viewers saw an ordinary person post one, they asked in DMs how it was made and users recommended the app, with no referral mechanism. Advice: make the app's output visible and distinctive rather than bolting on share cards or referral features later.

Before shifting to using creator content as ads, paid influencers to post on their own accounts in several countries, including Brazil.

With SKAdNetwork limiting attribution, the UA lead (running Apple Search Ads, Meta and TikTok on an ROI basis) gave up on predicted LTV and instead pushed the yearly plan so cohorts pay back in roughly 30 days.

Split everything by country and by platform (iOS/Android). Day to day, attribute trial starts to the same week's ad spend. For longer-term budget allocation, use blended cohorts: users acquired by a period's spend, per country and platform. RevenueCat is one source of new-user data. Planned next: a hybrid rule that re-counts users inactive for about 30 days as new, since heavy spend re-acquires past users.

Deliberately postponed CRM (lifecycle messaging) even past $1M MRR: the top five languages cover no more than two thirds of users, so localized CRM would take a lot of effort for an expected single-digit growth gain. Put the effort into monetization experiments and AI video features instead; plans to add some CRM later.