Trendspotting: YouTube Edition

As part of my new work consulting with clients on digital creative content development and scaling, video has obviously been the key focus, and YouTube longform content specifically.

At COURIER, in addition to our public-facing Epstein Cover-Up Database AI-powered project, we built internal AI-powered research tools to track the burgeoning progressive political creator space on YouTube (i.e., MeidasTouch, The Bulwark, Crooked, etc.). The purpose was two fold:

1. Identify what the trending topics were in the ecosystem in realtime
2. Just as important, identify the gaps – if every creator was following the bouncing ball over here, what were they missing over there.

(As an aside, we also prototyped similar tools tracking local government town councils livestreamed on YouTube across several key states, as a way for our local reporting teams to keep on top of secular trends happening across localities, such as the growing opposition to ICE warehouse detention centers.)

I wanted to experiment with pulling this tool out from behind the wall and see if it could be applied to non-political spheres. True Crime was an obvious place to start.

• It's a highly saturated space in the podcast world, with a high level of the same "chasing the big story" element I saw with political content.
• But it is a genre that is also inconsistently colonized on YouTube, with some of the biggest names in the genre still finding their way on the video platform.

So, using a set of simple code bases and off-the-shelf agents, I put together the True Crime Tracker here. The Tracker identifies the top ten most popular true crime videos/channels each week, analyzes and summarizes their content makeup in a unified Dossier. It then compares that to the last 8 months of similar programming to spot what topics are up, what are down, what are evergreen, and where the ball is going next in a Trend Intelligence tracker.

Take a look – comments and improvements welcome!