How Pushcroll grows

Every video ends with 'Comment "app" and I'll send you the link.' A Manychat automation replies to each 'app' comment with a DM ('Hey there, please click the button below to receive the link') containing a direct store link. A one-word comment is easy, and a wall of 'app' comments makes others join in. The main point is engagement: the comment flood signals Instagram to push the video; the direct link also removes the search step.

Modeled on Clear30's onboarding and Alex Hormozi's value equation (raise dream outcome and perceived likelihood, cut time delay and effort). Flow: 'Let us build your program', enter name, pick goals from many options (dream outcome), how much time you waste daily (pain), which apps are most addictive (later screens show that app), how it makes you feel (show the opposite feeling), 'What have you already tried?' followed by a screen explaining why each method (e.g. screen time limiters) fails, with tappable links to cited research, positioning the app as the best way to quit. Social proof screens sit between questions, and extra trust-building before the screen time and notification permission screens where many drop. A shorter version is being tested because the paywall rate is lower than they want.

Sent one templated Instagram DM (a hook, a quick app explanation, 'let me know if you're interested') about 2,000 times, then booked calls. Deal: 4 videos priced at their average views times a ~$2-3 CPM (100K avg views -> $200/video -> $800), flat retainer, no view bonus, 50% upfront and 50% after the last video. Signed 5-6 influencers, mostly calisthenics creators, plus a Chef Mac-style collab post copied from a format that worked for Quitter (shout-out in caption, co-posted).

Copied and slightly improved a competitor's format. Hook: filmed over the shoulder 'in the wild' at night or in the rain, the founder doing push-ups next to his phone on the pavement, caption and a female AI voice saying 'My bro is doing push-ups at 2:00 a.m. to scroll Reels' with a skull emoji. Within 2-3 seconds it cuts to him explaining and demoing the app ('every push-up I do gives me 1 minute I can use to scroll'), mentioning a secondary feature, and ending with a CTA. The more bizarre the setting, the more curiosity. Posted as an open ad rather than disguised. Repeated the same hook over and over; recommends having 3-4 formats in rotation because one fatigues, and studying competitors' hooks rather than copying 1:1.

Launch video ('After 14 days of working day and night...') told followers that everyone who installs in the first week (until a set date) gets lifetime free premium, including family and friends, and to share it before the offer ends. Android launched first with a 1-week window; the iOS launch reused the same video script with only a 48-hour window, requiring users to create an account and share the video, because they expected iOS users to pay more. After the window, new installs hit the paywall. Launch followers were funneled into a Discord.

Check the skip rate (share of viewers who swipe away in the first 3 seconds) in platform analytics. At ~30%, find a different format; iterate until it is around 18%. A low skip rate combined with good watch time is the strongest sign a video will spread. Spend about 80% of creative energy on the hook.

Tested a cinematic, story-style version of the push-up skit (character 'bound for eternity doing push-ups') that did not feel like an ad against the straight 'full ad mode' demo. Because viewers badly want to quit doomscrolling, they tolerate being sold to.

Because conversion was poor, A/B tested free trial length of 3 days vs 7 days, with ARPU as the north-star metric. Kept the 7-day trial and now works on making those 7 days show value (new features, retention).

Paywall variant with a UGC-style video (starts with a face, shows the push-up detection) followed by a notification bell visual, so users picture themselves using it; treats the onboarding and paywall like a short-form video with a hook. Tests are split into three cohorts (US, non-US under 18, non-US 18+) because the young European audience converts very differently.

Launched on Android first (Mario's existing code base) then iOS. Because their organic content reaches mostly European viewers, where Android dominates, Android users outnumber iOS about 3:1; they convert far less but make up the numbers. Advice: US-heavy audiences can go iOS-only; European ones need Android. Paywalls and pricing differ by platform, and Android users paid more.

A/B tested a weekly vs a monthly plan on the paywall, judged by ARPU.

Because price sensitivity is their main reason people don't subscribe, users who back out of the paywall see a second screen with a large discount ('80% off or something like that').

Before any product existed, posted a short video pitching the concept as if it were real: the doomscrolling pain point, then borrowed YouTube footage (googled 'AI push-up counter') of an AI skeleton overlay counting push-ups, explained as 'no push-ups, no scrolling', ending with 'Should I build this? Let me know in the comments. I post every day.' Simple language and a clear visual of the concept. It flopped on a new, unwarmed Instagram account but went viral on TikTok. Then found a technical co-founder on YC Co-Founder Matching and built a two-screen MVP in ~2 weeks. Recommends making one or several such videos and only building once one goes viral.

Post two design options (screens, logo versions, wireframes of a feature like the Duolingo-style 'journey' mode) in a Discord channel and let members vote (e.g. 102 votes on one choice), then redesign around the chosen version. Treat a suggestion seen ~20 times as a signal to build (the 'minute debt' feature, borrowing minutes to repay later, e.g. on a plane, was requested at least 50 times). Also use it to learn whether users care more about less phone time or more exercise. Works best for binary choices; keep your own vision rather than building every suggestion.

Built a Discord around the launch; the first and most active members came from Alejandro's TikTok followers who commented on the concept video. Channels for app decisions, bug reports and a self-improvement channel where users post progress (e.g. 300 push-ups, 8,000 push-ups). Early 'OG' members became moderators, and members send user-generated videos (e.g. push-ups in the mountains) the team can reuse.

To get content shown to US viewers from Europe they tried VPNs and US SIM cards, then remote-controlling an Android phone physically located in the US.

While short on money, paid Chris (ex-Quitter) for onboarding feedback; he told them to focus on onboarding and add social proof screens between questions. Also got feedback from other founders, and paid contractors for screen time bugs and design work. Combined with studying free resources (Superwall podcast, X, App Mafia).

During onboarding users can try the push-up detection (but cannot block apps yet) so they see it works and isn't a scam. They also answer 'When do you want to work out?' as a micro-commitment; an earlier version went further with a signed commitment ('From this day onward, I will...') with the user's name.

To reach the higher-ARPU US audience, hire UGC creators physically in the US with a US SIM card. Pay about $1,000/month plus a view-based incentive; avoid $400/month creators for the first 4-5 ('solid backbone'). Creators warm new accounts for 3-4 days (follow and engage, 3-4 niche posts like 'How many push-ups can you do?') before slowly promoting the app. Use UGC to scale a format already proven organically, not to explore; give creators less freedom than influencers. Tried AI-driven mass creation and preferred quality.

Push-up detection (Google MediaPipe, on-device for low latency) was first built side-on, as in most demos. They rewrote it to detect from the front and customized the skeleton overlay so the face stays visible, because front-facing reps with a visible skeleton are more appealing in creator and organic videos. Alejandro argues every consumer app needs a visual element for these platforms (Cal AI's calorie tracking, Quitter's streak).

Suggested by an ex-Superwall person and planned as a test: a dismissible paywall so users enter the app and get a ~15-minute scroll budget, then hit the paywall when they tap exercise detection to earn more minutes, at peak desire to keep scrolling.