An AI UGC video can nail the hook, the pacing and the product shot, and still lose people in the first two seconds because they can’t hear it. Plenty of viewers scroll with the sound low or off, and even the ones with sound on read faster than they listen. On-screen captions carry your selling line to both.
This guide covers what to do after the video is generated: what to put on screen, where TikTok’s buttons will cover it, which caption styles suit an ad, and how to reuse one winning video in other languages. It works whether you made the video with Oumomo’s Viral Remake or Link to Video, filmed it yourself, or got it from a creator.
Two kinds of “captions” on TikTok
The word means two different things on TikTok, and an ad needs both.
- The ad caption. The post text under your username. For Non-Spark ads, TikTok shows it in white, in a fixed font, with no clickable links, @ mentions or hashtags. Spark Ads take it from the organic post, up to four lines (TikTok Ads Help). Oumomo’s TikTok Caption & Hashtag Generator helps you write this part.
- On-screen captions. The words inside the video, timed to what the speaker says. That’s what this guide is about.
Both sit near the bottom of the screen, which is exactly why placement matters.
Why on-screen captions matter more in a UGC ad
- People read first. In a 2023 YouGov survey, 63% of US adults under 30 said they prefer subtitles when watching TV, even in their own language (YouGov). A phone on a bus or in an office is even less sound-friendly than a living room.
- The pitch is in the voice. In UGC, the problem, the price and the “here’s why it works” come from the speaker, not from a graphic. Without captions, a muted viewer sees someone talking about a product and gets no reason to tap.
- Everyone can follow. Viewers who are deaf or hard of hearing get the whole message, not just the visuals.
- You catch mistakes. AI voices and transcription still slip on brand names, sizes and prices. Captioning makes you read every word once more before you pay to promote it.
Step 1: Lock the audio before you caption
Captions follow the audio, so finish the voiceover first. If you wrote the script with Oumomo’s Viral Script Generator, play the final video against it: an AI voice sometimes changes a word, and the captions should match what is said, not what was planned.
Check three things before moving on:
- The product name is spelled the way it is on your listing.
- The price and offer match the product page exactly.
- Every claim is one you’re allowed to make.
Step 2: Generate word-timed captions
Typing captions by hand for a 30-second ad takes longer than making the ad. An AI caption tool transcribes the voice and times each word, and you only fix what it gets wrong.
- Export the finished video as a 9:16 MP4, ideally 1080 × 1920.
- Upload it to a caption tool and let it detect the language.
- Read every line against the audio. Fix brand names, numbers and slang.
- Split long lines so each one holds a single idea.
Keep one or two short lines on screen at a time, about 20 to 25 characters each. “Stains gone in 10 seconds” works; a full sentence wrapping over three lines doesn’t, because the viewer is still reading when the next shot starts.
Disclosure: I build Captions Generator, a free subtitle generator with no watermark, so this is the workflow I know best. Any tool that times captions to each word and lets you edit the text will do the job.
Step 3: Keep captions inside TikTok’s safe zone
TikTok draws its interface on top of your video: Following, For You and search at the top; the username, ad caption, sound and call-to-action button at the bottom; like, comment, save and share down the right. Captions placed in the “natural” spot near the bottom end up hidden under the ad caption.
For a 1080 × 1920 video, a safe area based on TikTok’s own safe-zone files is:
- Top: at least 130 px from the top edge.
- Bottom: at least 484 px from the bottom edge.
- Sides: at least 44 px from each edge.
- Right side, lower half: 140 px clear for the buttons.
TikTok notes that the exact area changes with ad caption length and extras such as anchors, so download the current files from TikTok Ads Help for a final check.
The practical rule: put captions just above the bottom third, below the speaker’s face, and never over the product.
To check a video in a few seconds, drop it into a TikTok and Reels safe zone checker and switch to the TikTok view. It shows what the app’s buttons would cover, and you can download a transparent PNG overlay for CapCut or Premiere.
Step 4: Pick a caption style that sells
The best ad captions are easy to read first and eye-catching second.
- Clean or boxed text. The most readable. Good for demos, reviews and tutorials.
- Word highlight or karaoke. The spoken word lights up as it’s said. Good for fast hooks, because the eye follows the voice.
- Keyword color. Color only the benefit or the offer, such as “no more frizz” or “50% off”. One colored word per line is plenty.
A few rules hold for every style: white or yellow text with a dark outline or box, big enough to read at arm’s length, and no animation that makes the viewer wait for the words.
Add a hook title for the first two seconds, such as “3 reasons my hair stopped breaking”, near the top but below TikTok’s top bar. It tells scrollers what the video is before the speaker gets there.
Step 5: Export a clean MP4
Burn the captions into the video, so they look the same on every phone. Export a 9:16 MP4 at 1080 × 1920 with no watermark: another tool’s logo in the corner of your ad costs trust you just paid for.
Keep the SRT file as well. It saves you from captioning again when you re-edit the video in CapCut or reuse it on YouTube Shorts.
Before the ad goes live, score it with Oumomo’s Video Quality Rating, then post or schedule it through TikTok Official Publishing.
Step 6: Translate the winners for new markets
When an ad works in one market, the cheapest test in the next one is the same video with translated captions.
- Translate the captions, then have a native speaker check the hook, the price and the offer.
- Re-split lines afterwards. German, French and Spanish usually run longer than English.
- Change prices and currency to the local version, not just the words.
Translated captions work best when viewers can still follow the visuals. If the whole pitch depends on the voice, make a new voiceover in the local language and caption that.
Pre-launch checklist
- Captions match the final audio word for word.
- Product name, price and offer are correct.
- No line sits under the username, ad caption, button or right-side icons.
- Captions don’t cover the face or the product.
- No more than two lines on screen at once.
- The export is 9:16, 1080 × 1920, with no watermark.
- The ad caption is written separately, with no links or hashtags on Non-Spark ads.
Frequently asked questions
Should I burn captions in or rely on TikTok’s auto captions?
Burn them in for ads. Captions inside the video look the same on every phone, and you control the words, the style and the position.
Where should captions go on a TikTok video?
Just above the bottom third, below the speaker’s face. On a 1080 × 1920 video, keep the last line at least 484 px from the bottom and clear of the buttons on the right.
How long should each caption line be?
Short. One idea per line, about 20 to 25 characters, and no more than two lines at once.
Can I caption an AI UGC video for free?
Yes. Most ads are under a minute, so free plans usually cover them. Look for one that exports in 1080p without a watermark.
Do captions change how TikTok delivers my ad?
TikTok doesn’t say that captions affect delivery. What they change is how much a viewer understands with the sound off, which shows up in watch time and clicks. Test it yourself: run the same ad with and without captions and compare.
Make every second readable
Generating the video is the fast part now. The few minutes you spend on captions decide whether a muted viewer gets your pitch or scrolls past it. Lock the audio, time the words, keep them in the safe zone, and reuse the winners in new languages.