Wrestle AI · B2C · AI wrestling coach app
Make polished, high-production video ads
Every attempt got near-zero revenue and very high CPA
Failedpaid-social
What they did
Went all out several times on 'really sick looking' visuals for ads.
What happened
Zero revenue and super high CPA every time; talking-head, stitch and raw demo ads did better.
- Effort
- several attempts ('every time')
In their words
The Superwall Podcast · George Lampropoulos - Meet the ex-wrestler making $41k/mo with AI apps“Ads that are very high quality suck, at least in our experience. Every time we've, like, went all out to make, like, a really sick looking visual, we've gotten zero, like, revenue off that. It would be like super high CPA.”
Related topics
Get tactics like this every Monday
New tactics from the week's founder interviews, each linked to where it was said.
More from this episode
- FightAI: Build an app you want yourself, then post about it to test demand~$1K revenue at pure margin; convinced him the idea was real
- FightAI: Spin off a version for the one sport users keep asking aboutWrestle AI passed 100K downloads; only wrestling app, so creators were cheap
- Wrestle AI: Make a VA running your Instagram DMs your first hireVA's DM through a mutual follower landed the niche's top creator as partner
- Wrestle AI: Make your niche's biggest creator a 50/50 co-owner of the appDrove launch and most growth; app passed 100K downloads
- Wrestle AI: Have a creator push an App Store pre-order before launch day$8K in week one; #19 in Sports on launch day
- Wrestle AI: Sponsor about 10 niche creators, then drop all but the bestPassed 100K downloads on creators alone; 'really profitable'
- Wrestle AI: Run the creator videos that explain your app most clearly as adsPaid ads just started; no results yet
- AbMax: Run ads where a creator reacts to a ridiculous video, then pitches the app~$60K back on ~$20K spend over 3 months; still the top-spend ad
- Wrestle AI: Run ads where a creator ranks the best apps and picks yoursFounder says ranking ads work as paid but flop organically; no numbers
- Brawler: Run a plain screen recording of the app in use as an ad$10 CPA on day one vs $40 break-even (3-4x ROAS)
- Brawler: Make many versions of each winning ad and raise spend slowlyFast spend increases raised CAC and erased gains; slow scaling grew revenue
- AbMax: Have more UGC creators copy your best creator's ad format
- AbMax: Work out your break-even cost per sale and judge every ad against itSimple to read while ads are the only channel
- Brawler: Pay a big-name creator expecting a loss, so your ads feel trustedRecouped 75% of $4.5K vs a 25-50% goal
- Brawler: Sponsor creators who teach the sport, not ones who post highlightsInfluencers bring ~25% of revenue; does well with the right people
- AbMax: Pay clippers for cheap views, even at a loss, to make ads convert betterJust started; aims for 10M views at ~$0.30 CPM
- Wrestle AI: In onboarding, ask about users' pain points, build their plan, then show the paywall
- Wrestle AI: Add a feature people use daily, like a calorie tracker, between big usesFounder says it helped retention 'a ton'; no numbers
- Wrestle AI: Try dropping the free trial
- Wrestle AI: Sell gift subscriptions to parents with ads at Christmas
- Wrestle AI: Require your sponsored creators to also post inside your app's feedJust launched; too early to tell
- Brawler: Pay a pro coach to record lessons users must perform to unlock the next
- Wrestle AI: Copy Duolingo's push notifications and tie them to what users didNo stats; they don't track which notifications get clicks
- Brawler: On a user's third app open, offer a phone call with a founder10-15 calls; found users never knew about features that took weeks
- Brawler: Show a short demo video the first time someone opens a tricky feature
- Ladder: Let people use the app free for 7 days before showing any paywall