RocketShip HQ - Mobile User Acquisition Agency

I just analyzed 100+ Facetune ads with AI. Here's how I'd 2x their growth as their creative strategist. What they get right: - Hairstyle reveal videos are their strongest format: quick cuts, TikTok pacing, "wrong hairstyle to right one." - They localize their winners instead of running one version: Spanish and English cuts of the same concept. - They're already building on what works: two of their three most recent test lanes rebuild that same hair reveal with a new creator or a new language. But here's what I'd do next. Their reviews keep bringing up background removal, skin smoothing, golden-hour lighting, and natural, subtle edits. Almost none of that shows up in the ads, which are nearly all hair. And even the hair ads show the transformation without ever showing how it happens inside the app. Here are 4 big opportunities to 2x their growth: Opportunity 1: Winner Extensions - Keep the "wrong hairstyle to right one" shell. - Swap the variable: new cuts, new colors, new creators, new languages. Opportunity 2: Adjacent Extensions - Same "wrong choice" mechanic, new category. - Wrong outfit, wrong makeup, wrong lighting, wrong nails before a party or a date. Opportunity 3: New Lanes From The Reviews - Background and object removal, skin smoothing, golden-hour lighting, natural and subtle results. - These are the most requested themes in their reviews and almost never tested in their ads. Opportunity 4: Emotional Specificity - Name the enemy: looking fake, looking over-filtered. - Paint the dream: looking glamorous, like a model, in a photo that's still you. - Prove the bridge: show the exact in-app moment that gets you there. The fix isn't more ads. It's different ads, built from what already exists. Drop FACETUNE in the comments and I'll send you the full video and deck. 👇

I read 1,025 Wispr Flow reviews before I looked at a single one of their ads. Every new-idea lane I'd recommend was already written in there by users. ↳ Hand surgery, tendonitis, ADHD, dyslexia: people with a medical reason they can't type ↳ Accuracy as proof: a thick accent, a non-Western name, and it still lands right ↳ Dictating in Hindi or Serbian instead of English ↳ Head to head against normal voice dictation None of those angles show up in the 126 ads I deep-dived. When people ask where new creative directions come from, this is the cheaper answer: the reviews are already sorted by what users cared about enough to write down. Full deconstruction in the link.

I just analyzed 335 Runna ads and 1,035 reviews with AI. Here's how I would 2x their growth as their creative strategist. Here's what they get right: ↳ A 4.29-star running app with 72% five-star reviews ↳ A proven copy formula: social proof + plan showcase + free trial ↳ Testing discipline: regular testing cadence. they kill 26% of what they ship That’s what’s gotten them to their level of scale as a TOP health & fitness app - and led to their acquisition by Strava. BUT they are sitting on a TON of opportunity - that can take their creative performance to the next level, and unlock their next level of growth. Here are 3 big opportunities to 2x their growth: Opportunity 1: Use emotionally resonant copy(steal from user reviews)** ↳ Use story-led copy drawn from real review quotes: - "Hated running, now I love it." - "I've never been a runner until I found the Runna app." - "Runna didn't just make me faster. It made me a runner." ↳ Run the narrative formats Runna isn't using: Social Proof, Ugly/Lo-Fi, Listicle ↳ Pair review quotes with lifestyle photos. Skip the plan selector. Opportunity 2: Tell the non-runner story Most ads target active runners training for an event. Reviews show a TON of users are non-runners who started running. BIG user segment worth targeting. Opportunity 3: Teach first, download later Most ads sell. Big opportunity to teach and gain trust. ↳ Pacing tips: '80% of runners set their easy pace too fast.' ↳ Training science: 'Why Tuesday is intervals, Thursday is tempo.' ↳ Form callouts: 'Your watch says you ran 5K. Your form says you ran 3K.' The fix isn't more ads. It's different ads, built from what already exists. The creative system, if fixed, can 2x their growth. Drop a comment saying RUNNA and I'll send you the full video and deck.

I just analyzed 100 Remini AI Photo Enhancer ads with AI. Here's how I'd 2x their growth as their creative strategist. What they get right: ↳ The swipe-reveal before/after is a universal mechanic that works across every language. ↳ Quality rescue is the foundation promise sitting under every ad in the library. ↳ The UGC creator template scales cleanly across markets and trends. But here's what I'd do next. One hook ("Are you team blonde or team brunette?") drove 16.3M reach over 161 days. It's the single biggest performer in the library. The same question architecture appears exactly once across the other 99 ads. The framework that grows this brand already exists inside it. It just needs to travel. Here are 3 big opportunities to 2x their growth: Opportunity 1: Make the question a system. ↳ "Team natural or team glow up?" for AI Flash. ↳ "Team grandma's wedding or team grandpa's army photo?" for restoration. ↳ Same community-vote architecture, applied to every product line. Opportunity 2: Name the moment, not the feature. ↳ 19 ads sell "restore old photos." Reviews tell a different story. ↳ One CA reviewer restored faded family pictures from the 80s and 90s. One GB reviewer prepped a slideshow for a family event. ↳ Same swipe mechanic, overlay names the anniversary, the memorial, the reunion. Opportunity 3: Bring the polish to Monday morning. ↳ AI Flash ads exist. All fashion. Zero work context. ↳ Reviews already mention resume photos and posting-ready headshots. ↳ Same template, new setting: the LinkedIn photo, the first-day Slack avatar, the resume shot taken on a Tuesday at lunch. The fix isn't more ads. It's different ads, built from what already exists. I put together a full analysis w/ recommendations for how I’d 2x their creative velocity. Drop 'REMINI' in the comments and I'll send you the full video and deck 👇

I just analyzed 53 Pok Pok ads with AI. Here's how I'd 2x their growth as their creative strategist. What they get right: ↳ Anti-stimulation hooks land. Parents identify with the calm-screen-time framing. ↳ Travel walkthroughs are tight: stress moment, device, app close-up, calm child, App Store close. ↳ One template can carry them. Top EU performer is one creative under 3 ad IDs, doing 1.66M cumulative reach. The system works. But the next layer is hiding in plain sight. Here are 3 big opportunities to 2x their growth: Opportunity 1: Teach the parent first. ↳ The hook lands. The ad stops there. ↳ Use the middle for a real parenting framework: overstimulation, dopamine logic, 4pm witching hour. ↳ Pok Pok shows up in the closing 25% as the embodiment, not the subject. Opportunity 2: Port the relief sequence beyond travel. ↳ Travel is 26 of 53 ads. ↳ Everyday-life lanes (restaurant, waiting room, dinner) are 3 of 53. ↳ Same calm-child arc. New setting. Empty lane. Opportunity 3: Lead with imagination. ↳ Parents already use imagination language constantly. The ads never lead with it. ↳ Take the strongest travel shell. Replace the generic play beat with a specific imagination invention. ↳ Show what she built. The fix isn't more ads. It's different ads, built from what already exists. The system is solid. Three extensions and the library 3x's its velocity without changing the brand. Drop Pok Pok in the comments and I'll send you the full video and deck 👇

If you're running a business on web and app, you may be missing out on hundreds of thousands every month, and here's why. There are three user types in every cross-platform business: - Installs on app, monetizes on web - Signs up on web, retains in the app - Shops on mobile, finishes on desktop Most ad accounts optimize for one. The other two pay the rent. Here's how the cost compounds. Three layers, only the first one shows up in this quarter's dashboard: - The bid trap: the network finds users who match the event you specified. - The model trap: the algorithm learns "your customer" from that event and amplifies the error at machine speed. - The allocation mirror: you shift budget toward whatever the algorithm reports as winning. Three years of this and you're not running a UA program. You're running a measurement error at scale. The math, on a $1M/month account: - 30%+ of users have their LTV event on a different surface than your conversion event - The LTV gap is typically 5x or more (subscription 5-20x, e-commerce 3-5x, games 10-50x) - Six figures of monthly spend chases the wrong cohort The customer-base drift over 36 months is the number nobody runs. The 5-minute test: - Pull the conversion event firing to each ad platform - Compare it to your CFO's definition of a paying customer - If they don't match, the algorithm has been training itself against the wrong user Mobile UA solved this years ago. Independent attribution. Structured postbacks. A signal infrastructure that let the algorithm learn from what actually mattered. Web never had it. Until today. AppsFlyer Web Performance Measurement went live this morning at MAU Vegas, bringing the mobile-grade signal standard to web. Their CPO wrote up what it actually unlocks - link in comments