I’ve been experimenting with YouTube over the past few months, and one thing that quickly became clear is that YouTube really is a search engine, just like Google. The difference is that it’s visual, but the same principle applies: the algorithm wants to give people the most relevant and engaging content for what they search.
Here are the two big areas I’ve noticed matter most:
1. Video Performance and Engagement This is the top factor. YouTube rewards videos that keep people watching. Some signals I’ve been paying attention to:
Watch time and retention: If viewers stay for most of the video, YouTube treats it as a strong sign of quality. If people drop off quickly, the video sinks.
Likes, comments, shares: These show YouTube that viewers found value.
Personalization: Results are tailored to each person’s watch history, so my videos show up more for people who already like similar content.
2. Video Optimization and Metadata Since YouTube cannot “watch” your video the way a person does, it still relies on the info we provide to categorize it. Here’s what I’ve been doing:
Title: Writing something compelling that includes the keywords someone would actually type in.
Description: Using the first couple of sentences to summarize clearly and naturally with keywords.
Tags: Mixing broader tags (“meal prep”) with specific ones (“vegan high protein meal prep”) so the system knows exactly where my video fits.
When both pieces come together, it seems to click: the algorithm knows what the video is about and sees that people are responding well. That combination is what has started giving my small channel some organic views.
I realize this is probably old news for some of you on here but I’m still learning and wanted to share this in case it helps anyone else who is just starting out.
@sarahggal I get what you’re saying and a lot of it is true, but I think people overrate tags. YouTube itself has said they barely matter anymore. The system can literally listen to every word you say in your video and scan every frame, so it knows way more than what we type into a box.
That said, I still throw in tags because it takes 30 seconds and doesn’t hurt. For me, the real growth always came from how viewers reacted. If people actually stick around and watch, the algorithm pushes the video. If they bounce, it doesn’t matter how perfect your keywords are.
Think of it like hosting a party. You can make the best flyer in the world, but if people walk out after five minutes, no one’s coming back. Same with YouTube.
Hey Jamal. The party analogy clicks with me! I've been burning the candle at both ends .. late nights and early mornings have become my normal lately, trying to squeeze in some time to learn more about YouTube!
But yes, you make a fair point. I’ve heard the same about tags being less important, and honestly I’ve noticed most of my watch time stats seem to line up more with what you’re saying. When people stick around, the video gets surfaced more. When they don’t, it fades.
I’m still experimenting because I don’t have much content up yet, so I figure it doesn’t hurt to put a little extra care into tags and descriptions just to cover my bases. But I do agree that how people respond is the real signal.
@sarahggal Great breakdown Sarah, thanks for sharing this. I’ve been focusing on this same part of the algorithm for my own channel and it really does make a difference. Even with just a handful of videos, I’ve managed to see consistent growth in a very specific niche just by keeping these basics in mind.
What I’ve noticed is that good optimization works like planting seeds. The views don’t all come overnight, but if the fundamentals are solid, the videos keep picking up more traffic over time. That steady flow of organic views has been more valuable for me than chasing quick spikes.
Appreciate you putting it out here in simple terms. It’s the kind of reminder that helps people not overcomplicate things when they’re starting out.
@sarahggal Very interesting to read, but I’m not totally sure how much of it is set in stone. I’ve always been a bit unsure with youtube. I suppose I should just dive in really and I’ve seen people say different things about how the algorithm works, and it feels like the “rules” are not always that clear.
I suppose I’m just wondering if YouTube changes things so often that what works today might not work a few months from now. Has anyone else noticed that?
I get the logic, but part of me still wonders how much each piece really matters in practice. Like, is watch time basically the whole game, or do things like likes and comments actually move the needle in a noticeable way?
The reason I ask is because I’ve played with with a couple of channels where I did everything by the book and still saw traffic tank after an update. It makes me question whether the “rules” we think we understand are the same rules the platform is really running on.
Another thing I’ve been chewing on is whether this mostly applies to search. A lot of people I know get more growth from suggested or browse, which feels like a different ballgame. If that’s true, then putting too much focus on search optimization might not get you as far as you hope.
And to your point Jamal, I hear you about YouTube being able to scan every frame and word, but I’m not sure it’s realistic to think that happens for every single video. The bandwidth alone would be crazy, especially at scale. I feel like metadata still has a role to play, even if it’s smaller than it used to be.
I guess I’m just a little jaded at this point. Feels like you can do all the right things and still get sideswiped. That’s the reality of playing in somebody else’s sandbox, ya know?