Challenge
A company launched an in-store cashback feature in Poland, but most people had no idea how in-store cashback works, and app installs in the market were getting pricey — $1.50+ CPI. They needed installs from people who’d actually use the app, not just download and delete.
Solution
I designed a one-level Facebook and Instagram campaign targeting Android users, with several ad sets at $15/day. Bid caps of $1, $1.5, $2, and $2.5 were tested to find the sweet spot for cost efficiency. All ads linked straight to the app store.
I scripted several short video ads with Claude, using my Proven Viral Hooks Collection to craft compelling openers for each ad. Using Eleven Labs, I generated voiceovers, then filmed my own shopping trips — groceries, supermarkets, fast food — and edited in Camtasia. Music came from Sumo, and I extracted compelling frames from the videos to design static creatives in Canva.
During testing and optimization, I saw that shopping videos significantly outperformed the static image ads. Underperformers were cut at $20 spend, and weak hooks were replaced with fresh 3-second openers. This constant refresh kept performance high.
Results
I hit 11,802 installs at $0.36 CPI with $4,238 total spend — beating the Poland market average by 4x.
- Reels placements across both platforms delivered 69% of results, with Facebook Reels leading at $0.33 CPI
- The 25-44 age group generated most of the action — 65% of total installs
- Female audiences consistently outperformed males by about 8%
- Video hooks hit 35-38% 3-second play rate, proving the openers grabbed attention
- Videos delivered $0.32 CPI vs images at $0.50 CPI — 36% more cost-efficient
The bid-cap testing kept costs tight, but the biggest lever was the creative: filmed shopping trips beat static images by 36%, and swapping in fresh 3-second hooks kept installs cheap as spend scaled. That cycle — many creatives, tested fast, doubling down on winners — is the system I run for brands now.