Latest / Video Marketing with Fexingo: Short Form, Long Form, and Video-First Strategy / The 1 Second Per Word Rule for Video Captions
Transcript
- Lucas: You know that moment when you're watching a video on mute — waiting for the bus, sneaking a scroll in a meeting — and the captions zip by so fast you miss half the point? That's not just annoying. That's a conversion killer. And there is a specific number that fixes it: one second per word. Luna: One second per word. So for a 30-word sentence, you'd need thirty seconds of screen time? Lucas: Exactly. And it sounds obvious when I say it out loud, but look at how many videos cram ten-word captions into a two-second window. The research — there's solid cognitive load research from the University of Nottingham from a few years ago — shows that the average viewer needs about one second to process each word of on-screen text when it's paired with audio. If the audio is faster, the captions should match the audio pace, but you still need to keep each caption chunk simple. Luna: But don't a lot of creators speed up captions deliberately for pacing? Like on TikTok, fast captions feel energetic. Lucas: They do, and there's a trade-off. Fast captions can signal energy, but if they outpace the viewer's ability to read, you lose the bottom third of your audience — people who process text a bit slower, non-native speakers, anyone with a mild attention deficit. The one-second-per-word rule is a floor, not a ceiling. You can go faster if the words are short and the context is clear, but once you cross into sub-half-second per word, you start bleeding retention. Luna: So you're saying the rule is more of a maximum speed limit than a target? Lucas: Exactly. Aim for one second per word for the densest sentences. For simple words like 'the' or 'and', you can compress a bit. But the mistake I see most often is creators treating captions as a transcription dump — they just shove the entire script onto the screen in two-line chunks without any timing consideration. That's where the cognitive load spikes. Luna: And it's not just about reading speed, right? There's also the split-attention effect — your brain has to split focus between the caption text and the visual action. Lucas: Bingo. That's the core of the research. When you have on-screen text plus moving images plus audio, your brain is doing triple duty. The one-second-per-word guideline gives your audience enough time to read the caption, glance back at the visual, and absorb the audio. If the caption is too fast, they'll skip one of those — usually the visual, which is often where the emotional or product message lives. Luna: I've noticed some big brands do this really well. Take Apple's product videos — their captions are always paced deliberately, almost slow. It feels premium. Lucas: That's no accident. Apple's video team has been applying this for years. They'll show a key phrase like 'A17 Pro chip' on screen for a full three seconds — that's three words, three seconds. Meanwhile, a fast-fashion brand might flash 'New collection drop now 40% off' in two seconds. Which one feels more trustworthy? Luna: The Apple one, for sure. But does the rule apply differently across platforms? On YouTube, people watch with sound more often. On TikTok, it's mostly mute. Lucas: Great question. On mute-first platforms like TikTok and Instagram Reels, captions are the primary narrative. So you actually have to be more careful with timing because there's no audio backup. If someone misses a caption on mute, they're lost. I'd argue the rule becomes even more important there. On YouTube, where audio is expected, you can ease up slightly — but only if the audio is clear and the captions are supplementary. Luna: So what's the practical takeaway? When we're editing a video, how do we enforce this rule without manually timing every single word? Lucas: There are tools now. Descript, for example, lets you adjust the caption speed globally. You can set a minimum duration per character or per word. I recommend starting with a 0.8-second floor per word and then manually slowing down key phrases — product names, calls to action, any data point you want viewers to remember. And always watch the video on mute with fresh eyes before publishing. Luna: I've also heard that you should keep each caption line to a maximum of two lines — and ideally just one line of text at a time. Lucas: Yes, that's related. The one-second rule is about duration per word; the two-line rule is about chunking. If you put four lines of text on screen, even with slow timing, the viewer has to scan a block, which takes extra processing time. So keep it to one line for short phrases, two lines max for longer sentences. And always align the text so it doesn't cover key visuals. Luna: Which brings up another point: placement. A lot of creators put captions at the bottom, but on TikTok the bottom is where the progress bar and buttons live. So you end up covering controls or getting clipped. Lucas: Right. That's a platform-specific nuance. On TikTok, I see more creators moving captions to the top or middle — just make sure they don't cover faces. Instagram's safety zone is different. But the timing rule stays the same regardless of position. Luna: So, if I'm a brand putting out a 30-second spot, how many words should I aim for in the captions? Lucas: Based on the one-second rule, you'd have about 30 total seconds of caption time, but remember that not every second needs a caption. You can have pauses. So a 30-second video might have 20 to 25 words across the entire script — not all at once. That's actually a good constraint. It forces you to simplify the message. Luna: That's surprisingly few words. Most 30-second ads have 60 to 80 words in the voiceover. But captions are different — they're more like the core message. Lucas: Exactly. The caption should be the irrefutable core. If you're running a video ad on LinkedIn, the viewer might have sound off by default. Your caption is your sales pitch. If it's too fast, you lose the deal. One second per word is a good benchmark for that final polish. Luna: It's interesting that this episode came on a day when I spent the morning captioning our client's new product demo. I actually timed it — I had a 12-word sentence that I had to fit into 4 seconds of screen time. I kept compressing the text instead of stretching the timing. Now I know better. Lucas: That's exactly the instinct we need to unlearn. The video editor's job is to serve the viewer's processing speed, not to cram in as many words as possible. This applies to lower-thirds in long-form videos too — not just social clips. It's a universal cognitive constraint. Luna: So, if this conversation has sparked something you've actually used in your own video editing, or even just made you think twice about the next caption you post — well, that's the kind of shift that keeps this show ad-free. If today was worth a coffee to you, there's a link that makes a real difference: buy me a coffee dot com slash fexingo. Lucas: Yeah, and it's genuinely appreciated. Not because we're asking for anything big — just that if the show saves you time or helps you avoid one bad caption decision, that's the value. And listener support is what keeps it independent. Luna: Alright, back to the rule. One last nuance I want to touch on: what about languages other than English? Does the one-second rule hold for, say, German or Japanese? Lucas: Great point. The research I cited was on English speakers. For languages with longer average word length, like German, you might need slightly more time per word — maybe 1.2 seconds. For character-based languages like Japanese or Chinese, the unit is per character, not per word. A good rule of thumb is to test it with native speakers. But the principle holds: give the brain enough time to decode the text and still watch the video. Luna: So there's no one-size-fits-all, but the one-second-per-word English baseline is a great starting point. Lucas: Exactly. Start there, test your retention metrics, and adjust. If you see a drop-off at the caption-heavy points, you know you're going too fast. The data is your friend. Luna: And on that note, I'm going to go re-caption that demo video from this morning. Lucas: Good. And if you're listening and you try it, let us know — does the retention improve? Because I'd bet it does.