We spent $180,000 on a brand video campaign last year that performed brilliantly in the US, UK, and Australia. Same video, same creative, same media buy. We localized it into German, Japanese, Korean, and Brazilian Portuguese. The localized versions collectively generated 4% of the engagement the English original did, across identical audience segments.
Not 40%. Four.
The localization budget was fine. The translators were competent. The subtitles were technically accurate. The problem was that someone had decided subtitling was sufficient for all four markets, and in three of them, it wasn't. Japanese audiences on YouTube overwhelmingly prefer dubbed content for anything over sixty seconds. Brazilian viewers on Instagram Reels scroll past subtitled content because the text overlays are too small on phone screens. Korean B2B decision-makers on LinkedIn treat videos without Korean voiceover as "not intended for us" and scroll past.
We fixed it. We went back and dubbed the Japanese and Korean versions, reformatted the Brazilian content for larger text, and re-ran the campaign. The corrected versions did 11x better. But we'd already burned through most of the media budget on the initial underperforming run. Total ROI: negative.
Why the subtitle-or-dub question isn't just about translation quality
Most discussions about video localization focus on linguistic accuracy. Is the translation faithful? Are the cultural references adapted? These matter, but they're table stakes. The real variable that determines whether a localized video actually performs is format — and format is determined by platform, audience, content type, and viewing context, not by what's easiest for the production team.
Platform behavior is the starting point. YouTube viewers in Japan expect voiceover. This isn't a preference — it's a behavioral norm so deeply embedded that subtitled content underperforms dubbed content by significant margins. The same holds in Germany, where dubbing has been dominant since the postwar period. Instagram Reels, TikTok, YouTube Shorts — short-form vertical video is consumed with sound off more than half the time across all markets. LinkedIn B2B viewers often have headphones on and will listen, so dubbing reduces cognitive load during vendor evaluation.
Content type changes the equation
A thirty-second product teaser doesn't have the same localization needs as a five-minute customer testimonial or a twenty-minute technical webinar. The shorter the content, the more tolerant audiences are of subtitles, because the cognitive investment is smaller.
A brand anthem video with soaring music and emotional narration loses almost all of its impact in subtitles. You can't convey tone, pacing, or emotional cadence through text on a screen. If your video's value is in how it makes people feel, dubbing is almost always the right call. Technical content is the opposite — viewers want accuracy and clarity, not emotional resonance, and subtitles let them pause and re-read. For long-form webinars, the practical middle ground is "critical segment dubbing": dub the intro, key talking points, transitions, and CTA, subtitle the rest.
Video localization UX: where most companies lose audiences
Subtitle design. Default placement at bottom center works for film but fails for corporate media, especially on mobile. Lower-third graphics, logos, or CTAs in the same space make subtitles unreadable. Move them. Integrate them into the frame design during production, not after.
Subtitle styling. White text with thin black outline is standard but vanishes on bright backgrounds and clashes with brand palettes. Custom styling — matching brand colors, appropriate shadows — costs almost nothing extra. Most companies never ask for it.
Reading speed. German translations typically need 10-15% more words than English for the same content. If your narrator speaks at 150 words per minute, your German subtitles either compress the message or demand superhuman reading speed. This is a UX problem needing a technical solution, not a translation problem.
The honest cost conversation
$8–$15/min/language Professional subtitling for a corporate video
$80–$200+/min/language Professional dubbing (single voice + studio + QA)
$120–$225 3-minute video subtitled into 5 languages
$2,000–$4,000+ Same video dubbed into 5 languages
These numbers explain why companies default to subtitling. But the question isn't whether subtitling is cheaper — it is — it's whether subtitling is cheaper when your localized video generates 4% of the original's engagement because the format was wrong. You didn't save money. You wasted your production budget plus your media spend.
At Artlangs Translation, we help companies make this decision based on platform data, audience behavior, and content type rather than assumptions or budget defaults. Professional subtitling and dubbing across 230+ languages, with format recommendations tailored to where your content actually lives and who's actually watching it.
