You’ve got a great brand video. Now you need it in German, French, and Spanish. Subtitles? Voiceover? Start again? Here’s how to decide.
Video localisation isn’t one thing. The right approach depends on what the video is, who it’s for, and how much the differences matter. A product demonstration has different needs than a CEO message. A training video has different needs than a social media clip. Choosing the right approach saves money where it can be saved and invests where investment matters.
Subtitles: cheap but often ignored
Subtitles are the simplest and cheapest localisation option. The original video stays intact. Text is added to convey what’s being said. Costs are low, turnaround is fast, and the original performances are preserved.
The problem is that subtitles don’t work for all content and all audiences.
Viewers often ignore subtitles, especially on social media where videos play without sound and attention is fleeting. Studies consistently show that subtitled content has lower engagement and retention than native-language audio. People read slower than they listen, subtitles obscure part of the image, and following text while watching video divides attention.
Subtitles work best when the visual content is the main event. A product demonstration where viewers are watching what happens on screen can work with subtitles because the voiceover is supplementary to the visual information. A talking-head interview where the speaker is the content works less well because viewers are trying to watch and read simultaneously.
Technical content often works with subtitles. B2B audiences watching product videos or technical explanations are motivated to understand the content and will tolerate the effort of reading. Consumer content, where attention is more casual, usually doesn’t perform as well with subtitles.
Some content types actively resist subtitles. Fast-paced editing makes subtitles hard to read. Multiple speakers make subtitle attribution confusing. Content with important on-screen text creates conflicts between existing text and subtitle text.
Voiceover: options and quality levels
Voiceover replaces the original audio with target-language narration. The visual content stays the same, but viewers hear the content in their language. This works better for casual viewing than subtitles and preserves the original production investment.
Basic voiceover – sometimes called “UN-style” because it resembles simultaneous interpretation at the United Nations – plays the translated narration over the original audio, which remains audible underneath. This signals clearly that the content has been localised rather than originally produced in the target language. It’s functional but feels like translation rather than native content.
Full voiceover replaces the original audio entirely with target-language narration recorded by professional voice artists. Done well, this sounds natural in the target language. Done poorly, the mismatch between the speaker on screen and the voice heard creates disconnection.
Lip-sync dubbing matches the translated dialogue to the original speaker’s lip movements. This is the highest-quality option for speaking-to-camera content but also the most expensive and complex. Translation needs to match not just meaning but also timing and mouth shapes, which constrains translation choices significantly.
The right voiceover approach depends on how visible the speaker is. Pure narration over product footage doesn’t need lip-sync. A CEO speaking directly to camera may need it to feel authentic – or may be better served by a different approach entirely.
When to start again
Sometimes the best localisation is no localisation at all. Creating original content for the target market may work better than adapting English content.
This is particularly relevant when the English content includes cultural references that don’t translate, when the speaking talent is the message, or when the market expectations for content differ significantly from UK norms.
A brand video built around British humour probably shouldn’t be localised for Germany. The humour won’t land, and what makes the video work in English becomes what makes it feel foreign in German. A German-market video with German humour – or no humour, since German brand communications often don’t use it – may perform better despite higher production costs.
Speaking talent matters too. A CEO message works because the CEO is speaking to employees or customers directly. Voiceover breaks that connection. For high-stakes content where the connection with the speaker matters, having the speaker record in the target language – even imperfectly – often outperforms professional voiceover over their image.
Market-specific content may be necessary when the story needs to be different. A case study featuring a UK customer doesn’t have the same impact in Germany as a case study featuring a German customer. Localisation can translate the words; it can’t make foreign examples resonate as local ones would.
Platform-specific requirements
Where the video will appear affects localisation choices.
Social media platforms favour native-language audio. Autoplay without sound means viewers decide whether to engage before hearing anything – so visual engagement matters most initially. But viewers who do engage often prefer audio in their language to subtitles. Short-form social content often benefits from full localisation despite higher costs because performance differences are substantial.
YouTube offers automatic captioning that covers basic subtitle needs for some content. This shifts the cost-benefit calculation – if YouTube will generate adequate subtitles, professional subtitle localisation adds less marginal value. But auto-generated subtitles aren’t good enough for technical content or content where accuracy matters.
Corporate websites and intranets have different requirements. Viewers are more patient, context is clearer, and polish may matter less than accuracy. Subtitles may work well for internal training videos where staff are motivated to learn regardless of format.
Trade show and presentation use requires different considerations. Videos played on loop at exhibition stands work without audio, so subtitles or on-screen text may be appropriate. Videos played during presentations need to work with the presentation context – which may mean different versions for different purposes.
Making the decision
The localisation approach should match the content’s purpose and the cost of getting it wrong.
For brand content where impression matters, invest in quality. Voiceover or full remake for key markets signals that you take those markets seriously. Cheap subtitles on brand content signal the opposite.
For technical and training content where comprehension matters, invest in clarity. Subtitles may work if viewers are motivated and content is visual. Voiceover may be necessary if audio explanation is integral to understanding.
For high-volume content where cost matters, standardise approach by content type. Not every video justifies full localisation. Having clear categories – brand videos get voiceover, product demos get subtitles, social clips get market-specific versions – simplifies decisions and manages costs.
For testing and optimisation, try different approaches. If you’re producing substantial video content for multiple markets, running tests to see which localisation approaches perform best in which contexts informs future investment.
At Bubbles, we help businesses make smart localisation choices for their video content. We understand that the goal isn’t to localise everything the same way – it’s to localise each piece in the way that makes it work for its target audience.
