optimized_poster.py replaces post_channel.py for channels where A/B testing is enabled. Instead of fixed hooks and title formats, it rotates through configurable variants and tracks performance data from YouTube Analytics to determine which variants are winning.

What Gets Tested

  • Hook overlay text: "STORY TIME" vs "WAIT FOR IT" vs "YOU WON'T BELIEVE THIS"
  • Outro call to action: "SUBSCRIBE FOR MORE" vs "WATCH NEXT" vs "LIKE IF THIS HAPPENED TO YOU"
  • Title format: Question format vs statement format vs emotional trigger format
Self-optimizing channels
After 10 Videos the Best Variant Wins Automatically
No spreadsheets. No manual analysis. The script reads YouTube Analytics and shifts to the best performer.
Get the Machine --

The Winner Selection Algorithm

After each variant has been used at least 5 times, the script calculates the average click-through rate and average view duration for each. The variant with the highest combined score is locked in as the default for the next 30 days. After 30 days, the test resets with new variants to account for audience fatigue.