Strategist answers · Commentary / reaction · 500K to 2M
Where do I start auditing 700 commentary uploads and Shorts at around 1.9M subs when growth has stalled and my four-person team keeps choosing topics from memories of our biggest hits?
For 1.5M-2M Commentary / reaction channels; small team (2-5); long-form + Shorts; stalled.
82% of this answer is quoted word for word from 6 YouTube strategists. Each quote links to its source. Last updated .
For a four-person commentary team choosing topics from memory, Daniel Faraday and Ed Lawrence question what past hits prove:
“Everything gets optimized by people who have studied what works. And what works - predictably - is what has already worked. So, why is this a problem for the creators?”
“Then you copy their strategy and it doesn't work, because you only saw half the picture. The channels that actually grow stop copying what worked for someone else and start figuring out what works for them.”
For an audit of 700 uploads, Dabi and Daniel Faraday both lean on data across many videos:
“Interestingly I actually did a massive informal study on this once when trying to build a views predictive model with several hundred K videos worth of metadata. So day 1 trajectory there was surprisingly super predictive of how the video would perform overall.”
“I compared every long-form upload since 2024 from 19 big operator channels with each channel’s normal: • Under 10 minutes: 0.62× normal views • Over 90 minutes: 1.25× normal views Short wasn’t the safe choice here. On these channels, it was most likely to underperform.”
For the Shorts in a mixed catalogue, Mario Joos gives a specific Analytics filter to start from:
“However, if you want to check this for yourself, go to Analytics, click on Advanced Mode, filter by Content Type (Shorts), filter by Publish Date (for example any short published from Jan 2025 to Jun 2025), set the data to Last 365 Days and take a look at the change happening around September. I think you can imagine why it took so long for us to find out where the issue began.”
Leroy ter Braak frames a per-video audit one way, while the 1of10 Podcast guest points to a channel-wide issue:
“Tell me how many views you think it deserved, what you think held it back, and I'll make a video breaking it down soon. I'll pick a few random ones I think are interesting! :)”
“what's the biggest mistakes you guys see when like you you guys have audited hundreds of channels, thousands of channels. Like what's What's the number one mistakes you keep seeing creators do? >> Um I think that one of the biggest issues that I keep seeing people run into is audience segmentation.”
For turning findings into new topics on a stalled channel, Dabi and Mario Joos offer different angles:
“Been thinking about ideation recently in a similar way. In the same way as that Wikipedia game where you gotta find page B starting from page A, maybe starting off from an okay idea and gradually tweaking it one bit at a time until it hits that “this is incredible” feeling.”
“But all of this is more short-term rather than long-term concerns. My other question, why do you need 3-4 weeks to restructure and why can the restructure not happen WHILE you're still creating content? There has to be other problems”
Export your catalogue with CreatorSide YouTube Scraper, then audit the complete video metadata CSV in a spreadsheet. Give your next topic meeting a record of past uploads to work from. YouTube Scraper
Every quote is copied word for word, credited to the person who said it and linked to the original post or video. Spotted a misquote or a missing credit? Tell us and we'll fix it.