A common piece of advice I see is to make your next upload 1% better. Here's how I actually do it.

I run a faceless gaming channel, and over the last two months (with 17 uploads), each video has had a small improvement from the last. Breaking it down video by video helps explain the process:

🎬 Video 1: First upload after a 5-year break. Focused on better audio quality, background music, screen effects, and cutting any dead air longer than 5-6 seconds.

🎬 Video 2: Cut the length by 70% - removed most failed attempts and testing. First video had 15% retention after 3 minutes, so this was a direct fix. Thumbnail got a text border and game-specific PNGs from the wiki.

🎬 Video 3: Watched the final edit all the way through (caught a vocal slip I missed in the first two). Adjusted thumbnail image sizes and got more comfortable creating a visual summary. Retention jumped to 55-70% for the first 30 seconds and 35-45% for the rest - stayed consistent after that.

🎬 Video 4: Learned to use blur effects more intentionally. Started cutting menu footage and moving the audio to better gameplay.

🎬 Video 5: Realized how many assets the game wiki had. Put more effort into thumbnail placement and avoided weird overlaps.

(3-week break for a holiday)

🎬 Video 6: Personal favorite. The challenge was simple but hadn't been done in the scene. First time dubbing dialogue in post to fix awkward lines. Thumbnail used the official game font instead of a custom one. Best performer of recent uploads.

🎬 Videos 7-10: Smaller tweaks - continued refining thumbnails, condensing cuts, cleaning up audio.

🎬 Video 11: First fully dubbed in post. Followed the general theme of the original commentary but without the stress of live gameplay.

🎬 Video 12: Used spreadsheet assets to show scores instead of just saying them or relying on in-game display.

🎬 Videos 13-16: General improvements to dialogue, cutting rambling and unnecessary bloat. Kept thumbnails consistent.

🎬 Video 17: First meme-style thumbnail trending in the community. Changed text display to include a border and slight drop shadow. Swapped out the blur effect for a cleaner one. Added 10-15 seconds of end screen for TV/desktop viewers to avoid the next video popping up too early.

Improving by 1% doesn't mean hunting for tiny optimizations. For me, it means watching each video after upload with fresh eyes and noticing my own habits as a viewer. Combining that with CTR, AVD, and impressions data helps build a clearer picture of what works for the channel. Over time, you start to see what needs to change.

Currently at 1.3k subscribers and 50,000 views over the last two months. 15 of 17 uploads have over 1k views, 13 over 2k. Not going viral, but building a consistent audience in the niche.

Happy to answer any questions in the comments.