July 25 @denjiyapss Check: Followers Slip as the Posting Gap Stretches

The July 18–25 check shows a small reversal in follower growth while likes hold steady. The public profile shows 1,557 followers, 17,600 total likes, and 9 videos. Compared with the July 18 baseline, that is -3 followers, no change in headline likes, and no change in video count. 1

There were 0 new public videos in the seven-day window. The account is still being carried by its back catalogue, but this week provides no new creative test to explain the movement.

The current public post list contains nine videos and reports no additional page. Its newest row is still the July 4 Airi rant/slime post, now at 364 plays, 30 likes, 0 comments, 0 shares, 1 save, and 36.4 seconds. 2

Because no video was posted in the window, there is no new weekly winner. The catalogue leader remains the April 18 slime/rant post, with 114,051 plays, 17,009 likes, 5,886 comments, 5,216 shares, and 4,822 saves. Its 57.5-second duration is also far longer than the July 4 post. 3

For the nine returned videos, rough engagement rate is calculated as (likes + comments + shares + saves) / plays. Saves use TikTok's public collect_count field. The rates below are comparison measures, not retention or private analytics. The underlying rows cover the April 16 to July 4 posts, including the five other rant/story-time or slime videos, two anime/game edits, and one trend clip. 4 5 6 7 8 9 10

The rant/story-time and slime group accounts for 119,108 of 121,008 catalogue plays, or about 98.4%. That headline needs a guardrail: remove the April 18 post and the group has 5,057 plays across five videos, averaging 1,011 plays per post. The format has the clearest evidence behind it, but the viral result is not a normal baseline.

Publish one controlled rant/slime video by August 1. Keep the experiment narrow: put the slime visual on screen immediately, use one clear opinion or story, and change one major variable from the July 4 attempt, such as the opening line. Hold off on a simultaneous edit or trend upload if the goal is to learn whether this lane is recovering.

Judge the result against two simple reference points:

If it clears 1,000 plays, repeat the lane once before adding another format. If it stays below that mark, run a second controlled variation before deciding that rant/slime has stopped working.

This issue covers the complete nine-video public list returned for @denjiyapss. The profile endpoint is used for the headline 17,600-like total; the post-list author snapshot shows 17,645 likes, so that separate snapshot is not used as the account headline. Private retention, watch time, traffic source, and audience demographic data were not available, so no conclusions are drawn from them. 1