content strategy / data analysis
I Analyzed 925 Finance TikToks. Only One Thing Separated the Winners.
Austin Starks scraped 925 videos from 24 finance hashtags on TikTok and had a video-understanding model watch the 41 most-viewed ones from start to finish. The obvious read is that the top of finance TikTok is teaching people about money. It is not. Fourteen of the 41 are wealth-lifestyle content with no instruction in them, three of the top ten contain no spoken words at all, and the second most-viewed video under finance hashtags is a political edit.
The part most analyses skip is the control group. A set picked as the top 41 by view count is selected on the outcome, so on its own it can only describe winners. Another 78 videos were drawn at random from the same scan and put through the identical pipeline. Across 39 comparisons of hook type, emotional driver, production quality, advice specificity, format and narrative structure, not one survived a Holm correction. The winners were no more specific and no better made than videos nobody watched.
One difference held. Content teaching investing, stocks or options is 25 of the 78 ordinary videos and zero of the 41 winners, at a permutation p of 0.00006, which survives a Bonferroni correction across every test that was run. It is not explained by audience size either: inside a matched follower band of 10,000 to 1,000,000 the gap widens rather than closing, and the hashtag data corroborates it independently at 2.7x.
Every figure here comes from the scan, and the full dataset and analysis code are in the repository behind the video.
Transcript
0:00I scraped nine hundred and twenty five finance videos and had a model watch the forty one biggest.
0:05Finance TikTok is not a finance category. It is a lifestyle category wearing finance hashtags.
0:11Three of the top ten have no words in them at all. One is someone failing to fan a stack of cash, and it did thirteen and a half million views.
0:20The problem is that every video I had looked at was already a winner, so I ran the same analysis on seventy eight ordinary ones. Almost nothing separated them. The winners were no more specific and no better made than the rest.
0:32Exactly one thing came back different. Thirty two percent of the ordinary videos were teaching you about investing, or stocks, or options. Among the forty one winners, that number was zero.
0:43The winning accounts were bigger, so I checked inside a matched follower band, and the gap got wider.
0:48So the subject is what separates them, and the videos teaching people about the market are the ones that lose.
0:55That is the entire category I make videos in, and I am going to keep making them anyway.
Join the conversation
Loading conversation…