Every creator knows the feeling: you finish a video, publish it, and immediately start staring at a blank page trying to figure out what comes next. You scroll through trending topics, check what other channels in your niche are doing, maybe ask in a Discord server. Meanwhile, sitting under your last ten uploads, there's a few thousand people who already told you in their own words exactly what they want to see.
The problem isn't that your audience is quiet. It's that comment sections are a terrible place to listen.
Comments are the most honest data you have
Think about what a comment actually is: someone watched your video, had a reaction strong enough to stop scrolling, and typed it out. Nobody does that for a survey. Nobody does that for a poll you post in your community tab. A comment is unprompted, unpaid, and critically it's about the thing you just made, not some hypothetical they're imagining for a focus group.
Compare that to the other ways creators try to figure out what to make next:
- Trending topics tell you what's popular in general, not what your audience specifically wants from you.
- Competitor research tells you what's already been made, which means you're always a step behind.
- Analytics dashboards tell you what happened (watch time, retention, click-through) but not why, and never what to do differently.
Comments are the only channel where your audience is actively, voluntarily telling you what they want in their own language, in the moment they're most engaged with your work. The signal is there. Most of it just never gets read.
Why creators miss it anyway
If comments are such good data, why doesn't every creator already have a comment-mining habit? A few honest reasons:
You only see the top of the pile. YouTube surfaces "top comments" by engagement, which skews toward jokes, pinned replies, and whatever got there first, not necessarily the comments with the most useful signal. A genuinely great video idea sitting in comment #340 might never surface.
Reading at scale doesn't scale. Skimming 40 comments on a small upload is easy. Skimming 4,000 on a video that did numbers is a part-time job. Most creators read the first page, feel like they've "checked the comments," and move on which means the majority of the actual data goes unread.
Repetition is invisible one comment at a time. The real signal usually isn't one great comment, it's the same request showing up 30 different ways across 30 different people. That pattern is completely invisible if you're reading comments one at a time in chronological order. You have to be counting, and nobody counts by hand.
What to actually look for
If you're going to read comments with intent instead of just skimming, here's what's worth paying attention to:
Direct requests, especially repeated ones. "Can you do a video on X" is the most obvious signal there is and if it shows up once, it's a data point; if it shows up a dozen times independently, it's a content gap you should probably fill next.
"I still don't get it" comments. These aren't complaints, they're a map. If multiple people say a specific part of your explanation didn't land, that's not a flaw to feel bad about it's a follow-up video that's already halfway scripted for you.
Comparisons. "This reminds me of—" or "have you seen [other creator]'s take on this" comments are pointing you toward collab ideas, response content, or an angle you haven't covered yet.
Timestamp reactions. When people call out a specific moment — a joke, a reveal, a strong opinion that's usually your best short-form clip, already identified for you by the people it resonated with.
Where the replies get heated. A thread where commenters are arguing with each other, not just with you, is a sign the topic has enough tension to sustain a full video addressing both sides. Controversy in the comments isn't always a bad sign sometimes it means you touched something your audience actually cares about.
The part that doesn't scale by hand
Here's the honest catch: everything above works great when you're looking at one video with a few hundred comments. It falls apart the moment a video actually performs, because now you're trying to manually track repeated requests, count how often a topic comes up, and separate real signal from noise across thousands of comments, with no comment thread saving your place.
That's the exact gap we built AudienzVibe to close. Instead of skimming the top comments and hoping you caught the pattern, it reads the full comment set on a video not just what YouTube chose to surface and turns it into a structured brief: what viewers are explicitly asking for, what topics keep coming up unprompted, and even short-form ideas pulled directly from the moments people reacted to. It's the same reading process described above, just done at a scale no one has time to do by hand.
You can see exactly what that looks like on a real video's report no signup required.
Your next video probably isn't a mystery. It's just buried on page fourteen of your comments, said by twelve different people who all worded it a little differently.