To answer these questions, I collected 120,000 public ranking records from the Chrome Web Store, built a prediction model, and identified the public signals most correlated with rankings. I previously shared some of the findings on X and Reddit, which sparked some discussion. This article is the full version — with a more detailed analysis, complete conclusions, and what I did with these findings afterwards.