How to pick faceless video topics that actually get views
Most faceless channels fail on topic choice, not on editing. A practical filter for picking angles that earn watch time.
The problem is rarely the edit
Faceless channels rarely die because the editing was bad. They die because nobody had a reason to watch video number two.
When you are not on camera, you lose the one advantage a personal channel has: people will watch a familiar face talk about nothing. Without a face, the topic has to carry the whole video. That is not a disadvantage — it just moves the work earlier, into choosing the angle.
The three-question filter
Before you generate anything, put the angle through this:
- Does it promise one specific thing? "Investing tips" promises nothing. "Why index funds beat most active traders" promises an answer.
- Would someone search this, or only scroll past it? Searchable angles compound; scrollbait dies in 48 hours.
- Can you say something that is not already the first result? A concrete number, a counter-example, or a named case. If the video repeats the consensus, viewers have no reason to stay.
An angle that fails any of the three tends to produce a video that looks fine and performs flat.
Format the promise, not the topic
The same topic can be a good video or a bad one depending on how it is framed:
| Weak framing | Strong framing |
|---|---|
| Sleep tips | Why you wake up at 3am (and how to stop) |
| Market news | What a 2% rate cut actually does to your rent |
| Productivity | The 90-minute rule that replaced my to-do list |
The right-hand column names a specific outcome, a specific person, or a specific number. That specificity is what the first three seconds of the video have to deliver on.
Write the first line before the script
If you cannot write a first line that makes someone stop scrolling, the topic is not ready. In practice this is a five-second test: say the opening sentence out loud. If it sounds like a Wikipedia summary, rewrite the angle.
This is also why generating a package is useful before committing to a topic: you get the hook and the scene split immediately, and you can judge the angle before spending an afternoon editing something nobody will watch.
Keep a topic bank, not a video bank
The channels that survive are the ones with a backlog of angles, not a backlog of finished videos. Ten angles that pass the filter are worth more than ten exported files, because the files expire and the angles keep producing.
A simple routine: once a week, spend thirty minutes writing only angles. Put every one through the three-question filter. Kill anything that fails. What survives becomes next week's production queue.