Update README.md

README.md

@@ -15,6 +15,11 @@ pipeline_tag: feature-extraction
 
 Tube Clip is a fine-tuned variant of CLIP Large (~1.7 GB) designed to analyze YouTube video thumbnails and titles to predict view performance classes.
 
+### Why?
+
+Content creators and marketers constantly rely on guesswork and A/B testing to figure out which combination of thumbnail imagery and video title will actually drive clicks. There is a massive disconnect between what _looks_ good and what the algorithm (and audience) rewards. 
+**So** I fine-tuned a CLIP model to evaluate the relationship between a thumbnail's visual elements and title with views classes. By analyzing 100,000 videos, this model removes the guesswork, providing an objective thumbnail score and class predictor based on proven, real-world data patterns.
+
 ## Model Overview
 
 * **Architecture:** CLIP (Vision + Text Transformer) fine-tuned on YouTube thumbnail and title distributions.