
Not in this cut: an offer. In a Sprint the offer (price, bundle, guarantee) is a test variable: a 1.5 s card swapped in before the end frame.
Beat boundaries sit on measured scene cuts (ffmpeg scene detection); ≈ marks one read from frames. The hook window is 0–1.5 s by definition. Click a beat to jump.
Muted. What still sells:
Most feed impressions play without sound. This is everything the cut says on screen, so it works muted.
- poppi
- 5g Sugar
- Strawberry Lemon
- Soda’s Back, but Better
Brand legible by ~1.0 s (frames sampled every 0.5 s)
Spec · Poppi · Beverage
Court side
Spec work · not commissioned by or affiliated with Poppi. Trademarks belong to their owners.Spec work · not commissioned by or affiliated with Poppi. Trademarks belong to their owners.
Tennis-ball fuzz at macro distance turns out to be a haircut; the can and its 5 g sugar claim ride the reveal.
- Brand
- Poppi
- Product
- Prebiotic soda can
- Format
- Paid-social cut
- Frame
- 9:16 · 10.1 s · with sound
- Brand on screen
- by ~1.0 s (frames sampled every 0.5 s)
- Tools
- Made with Higgsfield Marketing Studio (template remix) · edited in code
01 · Brief
Poppi sells “soda, but better” against legacy colas and a shelf of prebiotic look-alikes, so the claim has to feel like fun, not a label. This cut turns 5 g sugar into a visual pun that stops the scroll without sound and lands the can before the logo.
02 · Hook: the thumb-stop
Frame one is tennis-ball fuzz at macro distance. By second three it resolves into a haircut: the mystery holds the thumb, the reveal pays it off.
03 · Angle
Better-for-you, made playful: the can’s own claims (5 g sugar, Strawberry Lemon, “Soda’s back, but better”) arrive as hand-drawn doodles over a sky-blue court.
04 · Structure, beat by beat
- Hook · 0.0–1.5 s
Fuzzy yellow macro that reads as a tennis ball, then snaps to the can at 1.1 s: a curiosity gap.
- Problem / setup · 1.5–4.6 s
Can close-up with 5g Sugar, the can on a tennis-ball head, then “Soda’s Back, but Better” over the model.
- Proof / product · 4.6–8.9 s
The can floats in open sky; hand-drawn callouts repeat 5g Sugar · Strawberry Lemon · Soda’s Back, but Better.
- End card · 8.9–10.1 s
Chrome pink poppi wordmark over the model: brand lock-up as the end card.
Not in this cut: an offer. In a Sprint the offer (price, bundle, guarantee) is a test variable: a 1.5 s card swapped in before the end frame.
Timings from measured scene cuts; ≈ = read from frames. Turn on Ad anatomy to watch the beats on the player.
05 · What we'd test first
- A
Opener: the delayed fuzz reveal against a can-first frame one, judged on 3-second views and hold to second six.
- B
Claim order: the 5 g sugar doodle before the flavour name against after it, judged on hold to the end card.
06 · What a Sprint does with it
- Thumb-stop first. The first 1.5 s decides whether anyone sees the rest. We cut three openers from this master and let the account pick, instead of arguing about taste.
- One master, many cuts. This master becomes the 9:16, 4:5 and 1:1 set in two lengths, captioned and named for your ad account, so testing velocity is limited by budget, not by edits.
- Ahead of creative fatigue. New hooks and angles land before frequency climbs on the winner, so the account is never waiting on the next shoot.
- Judged on CAC, not on likes. Each cut has a hypothesis and a kill rule; the day-21 read-out says which one lowered your cost per acquisition and which one lost.
07 · Three frames

20 % · 00:02 
50 % · 00:05 
80 % · 00:08
Made with Higgsfield Marketing Studio (template remix) · edited in code
Brand owner? Email [email protected] and we'll take it down.