Text Rules for Reels, Stories and TikTok
Where each interface covers your video, how long a line should stay, and why 9:16 is not the same as 1080 × 1920 on every platform.
The same clip behaves differently on three platforms, not because the video changed but because the interface around it did. Designing for the smallest usable area and the largest overlay is the only approach that survives being cross-posted.
What each platform does to a 9:16 clip
- Instagram Reels: full-bleed vertical. Header at the top, caption and audio at the bottom, action buttons on the right.
- Stories: the same geometry plus a progress bar and a reply field, which sit lower than the Reels caption.
- TikTok: full-bleed vertical, a heavier bottom caption block, and a right-hand button column that reaches further into the frame.
- YouTube Shorts: the most generous of the four, with the overlay confined to the lower third.
Cross-posting is the normal case, not the exception, so design for the strictest of them: keep every element inside the middle band and away from the right edge.
First frame versus last frame
Feeds autoplay on mute, and the first frame is the thumbnail. A clip that opens on an empty background is a clip that looks blank in the grid. Start with the text already visible, or start it one frame into its animation so there is something to see.
The last frame matters for a different reason: the clip loops. If the final frame is a half-faded line, the loop point flickers. Ending on the fully visible line, held for a beat, makes the restart invisible.
Caption text versus on-screen text
They do different jobs and should not duplicate each other. On-screen text is a headline: three to six words that stop the scroll. The caption carries the sentence, the context and the call to action, and it is also where screen readers and search can find your words.
If a line is long enough that you are shrinking the type to fit, it belongs in the caption. If it is short enough to survive on a phone in one glance, it belongs on screen.
Duration and rhythm
Short-form feeds reward a fast first beat and a clear ending. For a text-only clip:
- 3 seconds for a single phrase, which loops cleanly and feels deliberate.
- 5 seconds when there is a second line appearing after the first.
- 8 seconds only for a sequence of three short beats, each held long enough to read.
Anything longer needs either movement in the background or a reason to exist. A static line on screen for twelve seconds is a still image with extra steps.
Compression is part of the design
Every platform re-encodes what you upload. Detail is the first casualty, which is why thin strokes, small outlined text and fine gradients inside letters degrade more than people expect. Two habits prevent it:
- Export at a decent bitrate. Around 8 Mbps for 1080p text is enough; the platform will compress once, not twice.
- Avoid pure white on pure black with hairline type. The edge ringing is the most visible artefact on a phone.
Accessibility, briefly
On-screen text is not read by screen readers. That is not a reason to avoid it, but it is a reason to keep the essential information in the caption as well. If the clip says "sale ends Friday" and nothing else does, a portion of your audience never receives that fact.
Burned-in captions are the usual compromise: readable by everyone, indexable by nobody. For a five-second text animation, the simplest fix is to repeat the words in the caption field.
A short pre-post checklist
- Is every word inside the middle 1400 pixels of the 1920-pixel height?
- Does the first frame already show text?
- Does the loop point cut back to the same image?
- Is the caption carrying the sentence the video is too short to say?
- Did you watch it once on mute, at phone size, before uploading?