You were doing fine. Now every video stops at 300. What changed?
The 300 band has a special sting: it mostly hits creators who had traction and lost it. The good news is that a regression has a diff, and diffs can be searched.
Why I wrote this
The 300-view searches read differently from the other low bands. They are past-tense searches: "was doing good, then...", "suddenly capped at...". This is the band of lost traction, and it hurts more than never-had-traction because it comes with a comparison and a suspicion that something was taken away.
The suspicion deserves a fair hearing, and a fair hearing is a diff: reach regressed at a specific point in time, so something changed at that point. There are only three places to look, and they go in order of likelihood: what you changed, what your audience changed, and what the platform changed.
The regression diff, in order
1. What you changed (the usual culprit, hardest to see)
Around the date reach fell: new format? new topic? new length? new posting rhythm, new editing style, a pivot toward "what worked" that quietly made every video the same? Creators change their content constantly while experiencing it as continuity. Line up your last twenty videos by date and look at them as a stranger; the diff is usually visible from outside and invisible from inside.
2. What your audience changed (the silent one)
A format that thrilled your audience in month one can be furniture by month four; trend formats decay fastest. The videos did not get worse, the novelty budget ran out, and the same first-audience behavior that once escalated your videos now reads as seen-it. The tell: engagement quality faded before reach did (fewer shares and comments per view in the weeks before the ceiling).
3. What the platform changed (real, but last)
Distribution systems do shift, and whole categories feel it at once. Before blaming this, check the account-status screens for restrictions (five minutes), and check whether peers in your niche regressed in the same window. If they did, adapt with the category. If they did not, the diff is closer to home. Platform-blame is the most comfortable diagnosis and, for a lone account's regression, the least commonly correct one.
The trap: flailing toward the past
The instinct at 300 is to resurrect whatever you were doing when numbers were good, and the graveyard of stalled accounts is full of creators re-making their old hits at higher production value. But if the audience's novelty budget for that format is spent, better executions of it spend nothing new. You end up polishing the exact thing the ceiling is made of.
The way out of a regression is forward: a deliberate portfolio of a few genuinely different next videos (different opener logic, different format, different promise), read not by their view counts alone but by which one earns disproportionate early engagement. You are hunting for your audience's next yes, and hunting requires spreading shots, not grouping them tighter on the old target.
Testing your way back, instead of guessing your way back
A regression hunt burns posts, and every burned post costs morale at exactly the moment morale is scarce. This is where pre-publish testing changes the economics: you can run the candidate directions past real tracked viewers first and see which opener actually holds strangers' eyes, before spending public posts on the search.
It also catches the subtler version of "what you changed": drift in the craft itself. When we eye-tracked 88 real short videos, 62% had a fixable attention leak the creator could not have seen in analytics, and the details were things creators do not notice accumulating: busier backgrounds over time, captions creeping onto faces, openers getting slower as confidence grows. A tracked panel reads your current videos the way the test audience does, and the diff between what they do now and what they did in your good months is usually right there in the first three seconds.
Find your way back with data, not nostalgia
Upload your next candidate videos to Jeena before posting them. Real viewers watch on their phones with the front camera on, and the reports show which direction actually holds strangers' attention: an attention heatmap, a visibility map, a wow-moments chart, and three concrete recommendations per video.
No "schedule a call." No sales rep. Upload, get your report.
Frequently asked
Why are my views suddenly stuck at 300?+
A regression to a low ceiling means something changed at a point in time, and there are three places to diff, in order of likelihood: what you changed (format, topic, rhythm, style, often invisibly), what your audience changed (novelty decay of a once-working format, visible as fading engagement quality before reach fell), and what the platform changed (check account-status screens, then check whether niche peers regressed in the same window). Single-account regressions usually trace to the first two.
Should I go back to the content that used to work?+
Usually not as a rescue plan. If the format's novelty budget is spent, higher-quality versions of it still spend nothing new, and you end up polishing the ceiling. Regressions resolve forward: test a few genuinely different directions, watch which one earns disproportionate early engagement (or run them past a tracked viewer panel before posting), and follow the strongest signal rather than the fondest memory.
Is a sudden 300-view cap a shadowban?+
Check it, then move on: account-status screens surface most real restrictions, and a strike or flagged content around the regression date is a legitimate lead. But if the account is clean and the drop coincides with changes in your content or your format's age, the mundane explanations fit better, and unlike a suspected shadowban, they are actionable this week.
What is Jeena?+
A testing service for short videos built on eye tracking. Real viewers watch your video on their phones, front camera on; Jeena logs gaze direction, blink rate, and eyebrow raises and collects a short survey. You receive the attention heatmap, the visibility map, the wow-moments chart, a perception summary, and three recommendations you can act on.
How does Jeena measure viewer attention?+
Each panel viewer calibrates their front camera once and watches normally. The platform then knows, frame by frame, where their gaze sat, when their eyebrows rose in surprise, and what they said in the survey afterward. That combination shows not just whether people watched, but what they were watching and how the video landed.
How much does it cost to test a video on Jeena?+
Roughly ten euros for a standard test. Check the pricing page for current rates.