[00:01] algorithm again, and there's some bad news and some good news. Let me explain. Before we go forward, we have to go back and do a brief history of the YouTube algorithm. From 2005 to 2012, the algorithm was keyword driven. So, that
[00:14] means that things like tags mattered cuz that was the metadata that supported and any YouTube guru that tells you that tags matter is a loser. Then, from 2012 to 2015, we entered into the attention driven algorithm. This was a big deal
[00:29] because before, people would spam keywords in their titles or their tags. They might put Justin Bieber in there to try to get more clicks, and it worked back then. But, what changed in 2012 was YouTube started to look at watch time.
[00:43] How long can your video actually hold someone's attention, not just get them to click. Then, the algorithm evolved again to the interest driven algorithm. YouTube, in a very precise way, learned who you are and what you like, and that
[00:58] influenced which videos were being recommended to you. But, most people again. But, before we talk about what the new algorithm prioritizes now, we have to look at two specific months this year where things changed. In January
[01:12] 2026, YouTube CEO published his annual letter on YouTube's own blog. He confirmed that AI is now behind recommendations. AI is getting involved in everything, dude. 1 month later, the browse feed was rebuilt around viewer
[01:26] watch history clusters. This is where things get really interesting as far as the opportunity on YouTube right now. Before this, the homepage was grouped in broad topics like cooking, tech, business. But, now these watch history
[01:40] clusters were much more specific. Now, instead of just being grouped as you're into business, it could be grouped as home service business owners researching pricing and hiring. That's a completely different level of precision. Which
[01:53] brings us to the new era of YouTube, 2026, which is the semantic algorithm. change is significant. Instead of YouTube just understanding your tags or YouTube just understanding your tags or even the keywords in your script, AI can
[02:08] now understand everything that is shown in your video and everything that's being said and what it means. Three things happened because of it. Number one, semantic IDs. YouTube reads what your video actually is, not what you
[02:23] titled it. Number two, we just learned about it, watch history clusters. YouTube groups viewers by specific, hyper-specific interest, not just broad hyper-specific interest, not just broad topics. And number three, satisfaction
[02:38] signals. It's no longer minutes matter most, did someone just watch the whole most, did someone just watch the whole video? It's did the right person love the video? The semantic algorithm is why
[02:50] a small specific channel can now reach exactly the right person without going exactly the right person without going viral.