How do I turn one video into multiple ad placements without re-shooting?

Validate a 9:16 version on TikTok or Reels first, then reframe the finished video to the other ratios instead of re-shooting. In Riffkit, reframing regenerates the same video at 4:5, 1:1, or 3:4 near frame-identical: same scenes, timing, and voice, only the framing shifts. One proven concept then covers Stories, Reels, and Feed with no new shoot and no new prompt.

How do I resize a video for Meta ads without cropping the subject out?

Don't crop. Cropping a 9:16 into 4:5 or 1:1 slices off heads, product, and captions because the composition was built for a tall frame. Reframing regenerates the video at the target aspect ratio so every shot is recomposed for the new shape. Run 4:5 in Facebook and Instagram Feed, 1:1 for carousel cards, and keep the original 9:16 for Stories and Reels.

Should I create every aspect ratio of an ad creative at once?

No. Most new creatives lose, so producing 3-4 ratios up front multiplies spend on losers. Test one 9:16 version on TikTok or Reels first, and generate the other ratios only for creatives that beat your account average. Since each ratio is its own video billed by the second, validate-first means the second round of spend goes only to winners, roughly 60% less production for the same coverage.

Can I reframe a vertical video to 16:9 for YouTube?

No. Going from vertical to landscape isn't a framing shift, it's a different composition: wider staging, more background, different blocking. Reframing covers the vertical family (3:4, 1:1, 4:5) from a 9:16 master; if you need 16:9 for YouTube in-stream, generate it as a separate video.

Turn One Winning Video Into Every Ad Placement (Without Re-Shooting or Re-Prompting)

Your best vertical video is doing one job when it could be doing five. Meta splits delivery across Feed, Stories, Reels, and carousel slots, and each placement has a native aspect ratio it rewards; TikTok and YouTube Shorts want vertical outright. The standard advice is to handle this up front: finish a creative, export it in every size, launch the whole family. That advice gets the order backwards. Most new ad creatives lose, so producing four aspect ratios of every concept means paying four times for every loser. The smarter workflow validates one 9:16 version first, then reframes only the winner into the other placements. Here's how to run it end to end.

Why "every size up front" quietly wastes your budget

Two structural reasons.

Creative testing is a search problem, and most candidates lose. Scaled DTC teams test roughly 5-15 new creatives a week, and only a fraction beat the account average. (If you're not sure what your weekly number should be, the benchmarks are in how many ad creatives you need per week.) When you produce a full ratio family for all of them, most of that production lands on videos that will never scale.

Generation has quality variance. Whether your creative comes from a shoot, an editor, or an AI pipeline, two takes on the same idea are never equally good, and you can't reliably tell which one is the winner by looking. The platform tells you, with spend. Making placement variants of an unvalidated video just multiplies an unknown.

Put those together and the principle falls out: placement coverage shouldn't be a re-creation job — it should be a framing variant of one concept the market has already voted for. Validation first, coverage second.

The validate-first workflow at a glance

Riff a proven concept into a 9:16 video for your product.
Test it on TikTok and/or Reels, the placements where 9:16 is native.
Reframe the winner into 4:5, 1:1, or 3:4 (same video, new framing).
Send each ratio to its native placement across Meta.
Repeat weekly, refreshing ahead of creative fatigue.

The rest of this playbook walks each step.

Step 1: Start from a video that already won

Don't start from a blank brief. Find a short video that is already winning in your niche: a TikTok in your category, a competitor-style ad format that keeps showing up, a UGC structure with proven retention. Paste the TikTok link into Riffkit (or upload a Reel or Short) and it extracts the formula (the hook, the pacing, the emotional beats) and generates a new 9:16 video that carries your product and story inside that structure. You riff the formula, not the footage, and the first version is ready in minutes, not weeks. Riff your first 9:16.

Starting from a proven structure matters for everything downstream: it raises the hit rate of your 9:16 tests, which means fewer rounds before you have a winner worth reframing.

Step 2: Validate in 9:16 before making anything else

9:16 is your cheapest honest signal. It's native on TikTok and Instagram Reels, both placements are built for fast feedback, and one vertical video covers both. Launch it and give it a clear bar: after enough impressions to trust the numbers, does it beat your account average on the metric you actually scale by (thumbstop rate, CTR, or CPA)?

Keep the window tight. Creative fatigue typically arrives within 7-20 days depending on the platform (TikTok burns fastest; see how fast TikTok fatigues creative), so a validation read should take days, not weeks. If it's clearly losing, kill it and move to the next concept. If it's winning, move to Step 3.

The discipline that makes this whole workflow work: produce no other aspect ratios until a video has earned them.

Step 3: Reframe the winner — same video, new framing

When a 9:16 wins, open it in Riffkit and reframe it to 4:5, 1:1, or 3:4. The output is near frame-identical: same scenes, same performance, same timing and voice — only the framing shifts.

And it's not a crop. Cropping takes a composition built for a tall frame and slices it into a square: heads clipped, product half out of frame, captions cut mid-word. Reframing regenerates the video at the target shape, so every shot is recomposed for the new frame instead of trimmed down to fit it.

Two mechanics worth knowing before you queue the batch:

Each ratio is its own generated video, billed by the second like any other. That's exactly why validating first pays: the second round of seconds goes only to creatives that already proved themselves.
Ratios you've already generated are skipped automatically. Reframe to 4:5 today and come back for 1:1 next week; Riffkit won't regenerate what already exists.

Step 4: Map each ratio to its native placement

With the family generated, place each version where it's native:

Ratio	Where it runs
9:16	TikTok, Instagram Reels and Stories, Facebook Stories, YouTube Shorts
4:5	Facebook Feed and Instagram Feed (the vertical-feed sweet spot)
1:1	Feed fallback, carousel cards
3:4	Portrait-feed surfaces; also the shape Instagram grid previews use

For the full placement-by-placement spec, including what each slot accepts and where the safe zones sit, see the video aspect ratio guide for TikTok, Reels, and Meta.

Because the reframes are near frame-identical, someone who watched your Reels ad and later hits the Feed version sees the same creative, properly framed, instead of a squashed or letterboxed version of it. One concept, visually consistent across every placement, which is exactly what you want for retargeting sequences.

There's also a volume angle. Meta's guidance for Advantage+ shopping campaigns is to feed the system 20-50 creatives. Reframes help you get there honestly (each validated concept becomes three or four placement-native videos), but keep the distinction straight: reframes multiply placements, new riffs multiply concepts. Advantage+ wants both, so keep the weekly concept testing running underneath.

The two exceptions: landscape and resolution

16:9 doesn't reframe from vertical. A landscape frame for YouTube in-stream is a genuinely different composition (wide staging, more background visible, different blocking), not a framing shift of a vertical master. If you need 16:9, generate it as its own video.

1080p is available at 1.5x the seconds of 720p. Treat it the way you treat extra ratios: an upgrade you buy for validated winners headed to placements where the sharpness earns its keep, not a default for test videos.

Both follow the workflow's one rule: spend the premium only after the concept has earned it.

The math: validate-first vs. full-family

Say you test 10 concepts a week and 2 beat your account average (a healthy hit rate).

Full family up front: 10 concepts x 4 ratios = 40 videos of production. The 8 losers took 24 extra videos (3 additional ratios each) down with them.
Validate-first: 10 test videos + 2 winners x 3 reframes = 16 videos for identical coverage where it counts.

That's 60% less production for the same placement footprint on every winner, and the advantage grows as your hit rate drops (new accounts, new niches, new offers). Each loser costs you one video, which is the correct price for an answer.

Run it as a weekly loop

The workflow compounds when it becomes a rhythm:

Early week: riff a batch of new 9:16 concepts from proven structures (5-15, depending on your spend tier).
Through the week: let TikTok and Reels vote. Kill losers without sentiment.
When a winner emerges: reframe it to 4:5 and 1:1, then rotate the family into Meta Feed placements.
Before day 7-20: refresh the winner with a new variation before fatigue takes it, and keep the concept pipeline full.

If an agent runs your marketing, the whole loop is scriptable: Riffkit ships a skill for Claude-based agents that can riff a source, watch the render, and trigger reframes from a terminal or a scheduled workflow.

Either way, the starting move is the same: pick one video that's already winning in your niche and riff your 9:16 version. Validate it, reframe it, and let one proven concept do the work of five placements.