You generated a short with Remakeit. The render finished, you posted it, and views stalled under 500. You tried more videos, more niches, more prompts. Same plateau every time. The output looks reasonable, the voice sounds okay, the captions are there — and yet the algorithm doesn't push the content. If this is the loop you're stuck in, the cause is almost always structural, not topical.
Here is the honest breakdown of why Remakeit outputs tend to underperform on TikTok, Reels, and Shorts, and the structural fixes that actually lift the curve.
Why Remakeit outputs often plateau
1. Hook is treated as the first sentence, not the first event. The generator writes a full script and the opener is rarely a real hook — it's more often a setup line that fails the 3-second cliff test. On short-form, this single defect is responsible for most of the early drop-off.
2. Visual rhythm is too steady. Outputs alternate visuals at a fixed cadence, often using similar style of images. After the first second, the eye predicts the rhythm and the brain disengages. Without an interrupt every 1–2 seconds, retention collapses.
3. Voiceover lacks emphasis. Default TTS presets are functional but rarely emphasize the punchlines. On short-form, a flat voice on a strong line loses to a charged voice on a weaker line. The energy in the voice is part of the hook.
4. Sentence-level captions, not word-by-word. Karaoke captions (one or two words highlighting at a time) outperform sentence-level captions by a meaningful margin. Tools that ship with sentence captions out of the box leave that retention on the table.
5. AI images often look the same across videos. Without strong style control, outputs from the same niche start looking like cousins. Viewers who saw three of your videos pre-judge the next one and swipe earlier. Visual uniqueness per video matters more than people think.
6. Limited control over the first 3 seconds. When the hook is weak, most regenerate flows rewrite the entire script and you can't surgically swap only the opener. You burn a credit, get a marginal variation, and waste the iteration.
7. Niche tone is generic by default. The script generator defaults to a neutral, informational voice regardless of the niche. Storytelling and opinion content require a different cadence, and the tool's defaults fight you on it.
Why prompt tweaks rarely fix it
It's tempting to assume that rewording the prompt will solve everything. In practice, the structural choices (hook design, visual cadence, voice realism, caption granularity) are baked into the tool's defaults. You can't prompt your way around defaults that don't match the format.
If a tool doesn't let you control those four levers directly, you will keep hitting the same retention ceiling no matter how you phrase the topic or how good the script is.
The structural fixes that work
Write your own opener. Don't let the generator handle the first sentence. Open the hook formula framework, pick a template, fill it with your specifics, and force it as the opening line of your script.
Add motion every beat. If the tool only outputs static stills, layer a zoom or pan in a secondary editor. Motion in the first frame is the single biggest pattern-interrupt lever.
Upgrade the voice. An ElevenLabs-class voice outperforms standard TTS noticeably on retention. If the tool doesn't offer one, render the script externally and remix it in.
Switch to word-by-word captions. Karaoke captions are the highest-impact tweak you can make if the tool ships with sentence-level subtitles. Render the video, add captions in a second tool if needed.
Cut the setup sentence. After the hook, jump straight to the tension or the reveal. Delete every "in this video", "let me explain", "today we'll cover" — those are retention killers.
A workflow some users switch to
A common pattern among creators who plateaued on Remakeit is to switch to a tool where the short-form structure is the default. Vexub was built specifically for TikTok, Reels, and Shorts — not retrofitted from a general video editor — so the structural defaults (hook-first opener, word-by-word captions, ElevenLabs-grade voices, motion-based visual rhythm) are tuned for the format out of the box.
What surprises most users is how little they end up changing. A lot of Vexub creators only edit the script. They type a topic, rewrite the first sentence with a hook formula, and post. Same default voice, same default caption preset, same default visual style. The reason the videos perform is that the structural defaults match what the algorithm rewards — the hook does the lifting, the caption animation holds the eye, the voice carries the emotion, and the visual rhythm alternates fast enough to prevent disengagement. The script edits become a small lever; the defaults do the rest.
How to know if it's the tool or your topic
If your 3-second retention is under 50 percent on every video regardless of topic, the cause is structural — switch to a tool tuned for short-form. If your 3-second retention is healthy (70 percent or higher) but completion stays low, the cause is the body of the video — your pacing or your reveal is the bottleneck, not the opener. Read the retention graph before you change tools.
The honest answer
Your Remakeit videos don't get views because the tool's defaults — setup-style openers, fixed-cadence visuals, flat TTS, sentence-level captions — are not the structure short-form rewards. The prompt isn't the problem. The structure is.
Rewrite the opener with a hook formula, add motion every beat, upgrade the voice, switch to word-by-word captions, and judge any tool on one metric only: 3-second retention. That number tells you whether the structural defaults are working for you or against you.
Read next: Why my video hook doesn't work · The complete hook formula framework · Why my InVideo videos don't get views · Why my Revid AI videos don't get views.
Create videos like this with AI
Script, voiceover, images and subtitles — automated in minutes.
