Why short-form video keeps getting shorter, and what that does to messaging

The compression of attention has been a media trend for two decades, but the pace of that compression has accelerated significantly with short-form video. When TikTok launched internationally, the maximum video length was 15 seconds. The platform has progressively increased its maximum length to accommodate longer content, but the content that performs — measured by completion rate and engagement — has trended shorter, not longer.

Data from 2025 platform analytics consistently shows that videos under 30 seconds outperform longer content on TikTok and Instagram Reels across most commercial categories. The exception is categories where educational content drives purchase consideration — some B2B categories, complex consumer purchases, and certain services — where longer-form content continues to drive meaningful engagement from high-intent audiences. But for the majority of product categories, the first five to six seconds are not an introduction — they are the entire opportunity to establish relevance.

What six-second relevance requires from creative strategy

Creating content that establishes relevance in six seconds requires a fundamental rethink of how brand communication is structured. Traditional advertising narrative — problem, solution, brand — is a sequence that takes time. In six seconds, you cannot present a problem and resolve it. You can state a specific, relatable situation that the audience instantly recognises as their own. The entire value of the hook is not capturing attention broadly, but capturing the specific attention of the person this product is for.

The practical implication for production is that the hook — the first three to five seconds — should be treated as a separate creative element from the rest of the video, developed independently, tested against multiple alternatives, and updated frequently. The body of the video matters significantly less than whether the hook succeeded in holding the viewer. Brands that treat video as a single creative unit rather than a modular structure are missing the optimisation opportunity that platform data makes available.