Consistency of attempts matters as much as any single video
Because performance varies video to video for reasons that are not always predictable, a channel that keeps posting genuine attempts finds what works faster than one that overthinks each video before publishing it. Volume, paired with real effort, is a legitimate growth strategy here. The recommendation system relies far more heavily on the video itself, the caption's words, the sound used, and how the first viewers react than on any external factor a creator can control in advance.
A recognizable format beats a broad topic commitment
A recurring series with a consistent structure, the same intro and the same beats each time, tends to build a following faster than a general commitment to a subject area. Viewers come to recognize the format itself, which gives them a reason to come back that a loosely related topic area does not.
The first few seconds decide almost everything
Since distribution depends so heavily on how the first viewers react, an opening that takes several seconds to get to the point gives viewers a ready reason to swipe away before the video has a chance to prove itself. Rework the first 2 or 3 seconds. Cutting straight to the most interesting moment rather than building up to it is one of the most effective edits available on almost any video.
Hashtags help the system understand a video; they don't place it
A small, specific set of hashtags that describe a video's topic helps categorization; broad, universally used tags carry close to no signal, since nearly every video uses them. Effort spent optimizing hashtag choice is better spent on the opening seconds and the content itself. Hashtags are a finishing touch, not a lever that moves reach on their own.