For the last 90 days I tracked every single link across our entire YouTube channel (just under 1.9M subscribers, ~900 videos) to see what’s actually driving business results — not just views. Because YouTube Analytics is great for showing you what’s performing on-platform (views, watch time, subs), but it only tells one side of the story. What I wanted to geek out on was the “flow on”: which videos are generating email opt-ins, which ones are driving affiliate clicks, and which topics are quietly moving the needle even if the views look average. To do it, I used Video Stats and then plugged the data into Claude via the MCP integration so the AI could analyze patterns I wouldn’t spot manually. The big surprise: more views does not mean more leads, more clicks, or more revenue. Some videos that I would’ve assumed “tanked” on YouTube (low views) were absolutely punching above their weight with massive conversion rates — and in a few cases the fix was obvious: YouTube was giving them impressions, but the CTR was low, so a thumbnail test could unlock way more of what was already working. The other big takeaway was intent. Tutorials and “how to” videos were lead machines, while comparison/best-of videos often attracted window-shoppers. Once you can see clicks, leads, and conversion rates per video, you can re-optimize descriptions, prioritize thumbnail updates, and build a content strategy that’s aligned with both the viewer and the business.