@ayoships's reel — transcript & breakdown

The speaker immediately grabs attention by revealing he used AI to reverse engineer a successful mobile app making significant daily revenue.

“So I used AI to reverse engineer a mobile app that's making over $100,000 a day.”

He sets up the main premise by claiming the app is 'gaming the ad algorithms' in an unprecedented way, creating a mystery.

“And it's gaming the ad algorithms in a way no other developer has talked about in their app.”

The video dives into the first specific method: how a single $40 purchase sends five signals to ad trackers, exploiting how ads optimize by signals, not dollars.

“When someone makes a $40 purchase, the code actually sends five purchase signals to its ad trackers for that one sale. And since ads are optimized by signals instead of dollars, that one big spender counts five times.”

He further explains additional signal-sending mechanisms for specific user behaviors like early or repeated purchases, highlighting a complex tracking strategy.

“But that's just the tip of the iceberg. It also separately sends signals for people who pay within a day of signing up, come back as a second day as paying users, or make a second purchase.”

Connecting these tactics to the app's large number of ads, he reinforces why the app is so profitable.

“Combine this with their 9,000 distinct ads, it's no wonder they're making so much money.”

The video reveals a second strategy: manipulating app store reviews by filtering negative feedback before showing the official App Store prompt.

“And the game also chooses who to ask for a public review. It first shows its own rating screen with five stars already selected. So if you keep it at five and submit, then it opens the App Store modal. But pick anything lower than five stars, it just thanks you and keeps the score to itself.”

He quickly mentions that Apple's rules don't permit this review tactic but advises copying the timing of review requests (after a win) for better results.

“Now Apple actually does not allow this pattern, but the timing is worth copying since it only asks after a win.”

The speaker directly engages the audience by asking for suggestions for his next app deconstruction, encouraging interaction.

“Tell me which app you want me to decompile next in the comments below.”