Guide
Two frames, one feed. The split-screen ad — a reacting face on top, the thing they're reacting to below — has become a default format in paid social for a simple reason: it gives a restless thumb two reasons to stay instead of one.
Split-screen video stacks two panels in one frame — usually top and bottom for vertical placements, sometimes side by side. The best-known variant is the reaction ad: a creator films themselves responding to footage, gameplay, a product demo, or a screen recording, and that footage runs in the second panel. The reacting human gives the ad a face and an emotion; the second panel carries the information. It reads as native content because it looks like the duets, stitches, and reaction videos people already scroll for.
A single talking-head shot asks the viewer to care about one thing. A split frame offers two: the reaction and the payoff. The face supplies social cues — surprise, delight, skepticism — that tell the brain something is worth attending to, while the lower panel keeps a second narrative moving. There's almost always motion somewhere on screen, which fights the flat, swipe-past feeling that kills static-looking ads. Done well, the two panels reinforce each other: the face tells you how to feel about what the second frame is showing.
The format isn't free. Divide the frame badly and you don't double attention — you split it. If both panels demand reading at once, or the reaction is generic filler that adds nothing, the viewer does the math and leaves. Tiny panels lose legibility on a phone, and on-screen text crammed into half a vertical frame becomes unreadable. The reaction also has to be earned: a face pretending to be amazed at second one, before anything has happened, reads as fake and gets swiped. The format amplifies a good idea and exposes a hollow one.
Splitting the screen doesn't exempt you from the first rule of paid social: the opening seconds decide everything. With two panels, you have two openings to land at once — the reaction has to register as genuine and the second frame has to promise a payoff, both inside the first three seconds. That's a harder hook to write, not an easier one. The same standard from hook rate applies — and if anything, a cluttered dual frame raises the bar for clarity in those first moments.
The honest question with any format experiment is whether it helped this particular ad — not whether it works in general. Rather than spend impressions to find out, run the finished cut through PreTestAds: predicted engagement scoring shows whether attention holds across the opening and survives into the CTA window, or whether the divided frame quietly bled viewers. Score a split-screen version against a single-frame version of the same ad and let the prediction tell you which layout earned its place. It's the same compare-before-you-spend logic behind all ad pre-testing.
Upload your dual-frame ad and check predicted engagement — first analysis free, no credit card.
Test Your AdA split-screen ad divides the frame into two stacked or side-by-side panels — often a person reacting on top and the product, gameplay, or footage they're reacting to below. Reaction ads are the most common version: a creator's face responding to content keeps a human anchor on screen while a second story plays out beside it.
Two frames give the eye somewhere to go. A reacting face supplies emotion and a reason to keep watching, while the second panel delivers the substance. The format mimics native feed content people already watch, and the constant low-level motion across both panels resists the scroll better than a single static-feeling shot.
Yes. Upload the finished dual-frame cut and predicted engagement scoring shows whether the layout holds attention through the first seconds and into the CTA — or whether the divided frame splits attention instead of doubling it. That tells you if the format helped this specific ad before you pay for impressions to find out.