The solution
In 2013, indie developers Alex Schwartz (Jack Lumber) and Ziba Scott (Girls Like Robots) had awards, publisher interest and good expo showings, but no income. At the 2013 Global Game Jam they bought a $15 3D slot-machine asset from the Unity Store and spent the jam building a system that generated just enough variation — change the title, change the one jackpot image, swap in a scrolling themed background — to publish each skin as its own free app.
The pipeline was end to end: feed in a keyword, and a Unity script assembled the game, complete with a custom song that used text-to-speech to sing the theme word in a Google Translate monotone every time the player won ('It'd say the name of your game in the music'). Image scraping tripped Google's rate limits, so they switched to Bing Image Search. A Selenium script then mimicked a human through Google Play's upload flow — clicking checkboxes, picking countries, agreeing to terms — while minding the 15-apps-per-day cap, ultimately uploading roughly 1,500 apps from a single account.
They attached the Playhaven ad network, never taking money from users, and walked away for two months. The apps pulled roughly $211 a day in ad revenue, with 27% of players clicking ads. Schwartz's theory: 'the quality was so low in our shit that the ads were a portal to a better world' — people who downloaded a bad casino game clicked through to genuinely better casino games. Ratings sat around four stars; one review of '3D Bowling Slots' complained the slots didn't have much to do with bowling. Total take: about $50,000 over a couple of years.
What it achieved
The terrible slots weren't the product — the ads were. 27% of players clicked away to better casino games, paying about $211 a day for the privilege of leaving.
Why it worked
They stopped competing on quality and automated the race to the bottom instead — a $15 asset became over a thousand apps no human hands touched.
App names followed a safe template — '3D' plus an adjective plus an animal, place or country plus 'free' — dodging trademark takedowns so the loop never needed a human.
Casino-related ad keywords turned each bad app into a funnel: players clicked away to better slots, and the exit paid better than the game.
What can be applied
Automation changes what 'a product' means: when creation is free you can serve the bottom of the market profitably — and find the bottom monetizes in ways the top never will.
Aftermath
The duo grossed about $50,000 over a couple of years from an experiment they treated as a dystopian joke, then presented it at GDC 2019 as a wake-up call about where app-store economics actually pay. They maintained they never wanted to take real money from the people who downloaded the apps.
FOLLOW THE EVIDENCE