How video ads earn the watch

Lengths that hold#

Prospecting video should land in 15 seconds or under: one idea, one proof, one action. Longer cuts belong in retargeting, where warm viewers grant attention cold ones never will. There is no universal ideal length, only the shortest cut that keeps the payoff.

Build modular masters: a 30-second story that cuts cleanly to 15 and 6 seconds without reshoots. Our hooks guide times the opening beat by beat.

Subtitle every master at source: clean dialogue tracks make captions, translations and re-edits trivial for the next two years. Archive the project file with its results attached. Caption discipline starts in the edit, not in export.

Ratios per surface#

9:16 full-bleed (1080 by 1920 or better) for Reels and Stories, 4:5 for Feed, 1:1 where square templates persist. Respect the Reels safe area: roughly 14 per cent top, 35 per cent bottom and 6 per cent sides kept clear of key content, with disclaimers clear of the bottom 40 per cent.

Meta own video specs list every placement crop. Design 9:16 masters first, then reframe outward; pillarboxed landscape in Reels reads as an afterthought. The hub places video in the family.

Export textless masters alongside captioned cuts so future edits, translations and ratio reframes never fight burned-in words. Archive project files with performance data attached; the opening that won in spring deserves a rematch in autumn. Diversification compounds when winners stay retrievable.

Sound-off by default#

A large share of Feed and Reels viewing starts muted, so captions carry the story: burn them in or upload en_GB SRT, size them for phones, and never let the punchline live in audio alone. Music and voiceover are enhancement, never load-bearing structure.

Test every cut muted before shipping: if the muted version confuses, the captioned version merely decorates confusion. Captions guide has the legibility rules.

Hooks in three seconds#

The opening must show the brand, the problem and the payoff direction before the thumb moves. Pattern interrupts earn one extra second; promises earn the watch. State the payoff plainly rather than teasing it.

Cut three openings per concept and test hooks before grading: openings decide, endings merely confirm. Try the creative checker on the hub before spending, and read Meta own diversification research for the evidence.

Production that pays#

Creator and customer-shot footage consistently beats studio gloss for prospecting trust per pound spent. Spend production budget on more concepts, not more polish: four rough openings out-pull one graded hero.

Pay creators for concepts, not just deliverables: the angle they find in your product often outperforms the brief you wrote. Renew usage rights before winners expire, or success gets expensive at renegotiation time. Partnership terms belong in writing from day one.

Keep a standing brief with two or three creators so new angles arrive monthly without a shoot cycle. UGC versus studio weighs the evidence honestly.

Reading video results#

Read 3-second views and thumb-stop rate for the hook, completion rate for the story, outbound CTR for the offer. A video with strong hooks and weak CTR has an offer problem, not a creative one.

Compare against the image control on CPA, not against other videos on views. Ask our team for a video audit when the numbers disagree with your instincts.

Log hooks tested with their thumb-stop rates in a shared sheet; the library teaches every later brief what this audience pauses for. Retire hooks below the account median ruthlessly each month. Fatigue curves show when winners turn.

Frequently asked questions#