Do UGC videos actually need captions?

Yes — and the sooner you treat them as part of the deliverable instead of decoration, the fewer revision emails you'll get. Feeds autoplay muted, a big share of viewers never unmute, and captions are basic accessibility for deaf and hard-of-hearing viewers. An uncaptioned video is silent for much of its audience.

Two kinds of text do two different jobs here:

  • Captions transcribe what you say, phrase by phrase, so the video works with the sound off.
  • Text overlays are headlines — the written hook in the first second, the "3 things I noticed" label, the price answer at the midpoint.

The overlay is retention equipment: many viewers decide to stay off the written hook before your first spoken line lands. (For the hook lines themselves, see how to write UGC hooks.)

So the working rule: if there's talking, caption it; if there's a hook, write it on screen. Brands see it in their ad data too — which is why captions now show up in briefs as a requirement, not a suggestion.

What caption style do brands want — and what reads amateur?

Clean and native. White sans-serif, medium weight, a thin dark stroke or soft shadow, two to four words on screen at a time, one consistent position. It's the look the platforms generate themselves — which is exactly why it reads as native rather than "made in an app."

Amateur is more specific than people think. The tells:

  • An uncorrected auto-caption error on the brand or product name. Nothing says "I never watched this back" faster.
  • The rainbow karaoke template — bouncing words, auto-inserted emojis, a highlight color per syllable. Fun on your own page; on a brand's ad account it dates the creative instantly.
  • Three fonts doing three jobs.
  • Captions drifting out of sync with your voice.
  • Text parked on top of the product during the demo.

One honest nuance: the punchy highlighted-word style does perform for some organic creators — if the brand's whole feed looks like that, match it. When the brief is silent, though, default to clean and let your delivery carry the personality. I've seen brands ask for a re-caption purely over style; I've never once been asked to make captions more decorated.

Where are the safe zones on TikTok and Reels?

Keep every word in a center band — roughly the middle 60 percent of the frame vertically, pulled in from the right edge. Platform UI eats the rest.

The specifics on a 9:16 frame:

  • TikTok: the icon rail covers roughly the right 15 percent, the username and caption sit over the bottom quarter, and the top tenth holds the search bar and feed tabs.
  • Reels: greedier at the bottom — Meta's own ad specs say to keep text out of the top 14 percent and the bottom 35 percent, because the username, caption, and audio row stack taller than TikTok's.
  • The feed crop: a Reel surfaced in the main Instagram feed is cropped toward 4:5, so text near the very top or bottom of your 9:16 frame can vanish entirely.

My placement rule: captions sit centered, a touch below the vertical middle; the hook overlay lives in the upper-middle third. Then check the boring way — preview on an actual phone, thumb over where the UI lands. Thirty seconds of squinting beats a revision email.

What's the fastest captioning workflow that doesn't look templated?

It's 11pm, the edit is done, and captioning feels like the last chore between you and delivery. Good news: it's about five minutes — auto-captions, one saved style, one proofread pass.

In CapCut:

  1. Run Auto Captions. Text → Auto Captions transcribes the whole video in under a minute. Accuracy is genuinely good now — call it 90 percent.
  2. Style once, save a preset. White, medium-weight sans-serif, thin dark stroke, no animation — or the subtlest pop, nothing bouncier. Apply to all, save it, and every future video is one tap. The trending word-bounce, auto-emoji templates are exactly what "templated" looks like — the saved preset is how you skip them.
  3. Proofread the 10 percent. Auto-caption errors cluster exactly where they hurt most — brand names, product terms, numbers. The serum's name arrives as three random words; "SPF 50" comes out "spf fifty." Read every line against the audio and match the brand's own casing. A misspelled brand name in the brand's own ad is the fastest way to look careless.
  4. Position, then phone-check. Nudge the block into the safe zone and watch once at arm's length. Any line you can't finish reading, split.

The edit around this pass lives in how to edit UGC videos — captions are the last clean coat.

Should you deliver captions baked in, or a clean version too?

Both. When you're delivering files rather than posting from your own account, send the captioned cut and a clean master — the same edit with no captions and no overlays.

The reason is the brand's ad tooling. Teams re-caption in their own editors or the ad platforms to translate the video, restyle text in brand fonts, or reposition it per placement — Stories, Reels, and in-feed each cover different parts of the frame. A baked-in caption block they can't move becomes a revision request; a clean master becomes "perfect, thanks." More and more briefs ask for the pair outright.

When yours doesn't say, one line in the delivery message settles it: "Captioned and clean versions attached — happy to match a caption style if you have a preference." That line has saved me more revision rounds than any editing trick.

Two habits finish the job: name the files so nobody guesses — brand-hookA-captioned.mp4, brand-hookA-clean.mp4 — and if a brand asks for an SRT, that's just your captions as a text file with timestamps; desktop editors export one in a click. The wider handoff checklist is in how to deliver UGC videos.

Caption clean, save the preset, send the pair. It's ten quiet minutes that mark you as someone who's done this before — and that's who gets rebooked.

FAQ

Do captions help TikTok SEO? Yes. TikTok indexes spoken audio, on-screen text, and captions for search, so naturally naming the product and what it does helps the video get found. Write for humans first — one natural mention beats stuffing.

Should captions be word-by-word or full phrases? Short phrases, two to four words on screen at a time. Word-by-word pop is a style choice, not a default — use it only when the brief or the brand's feed clearly calls for it.

What about videos with no talking? Nothing to transcribe, so no captions — but the written hook still earns its place. Label the file "no VO" so the brand knows the silence is intentional.

What font should UGC captions use? A clean geometric sans-serif — the platform-default look. Skip the brand's logo font unless they send you the file and ask; close-but-wrong brand type looks worse than neutral.