Meta won't run a playable ad on its own. It needs a separately uploaded lead-in video, and that video is the first thing the viewer sees. TikTok requires at least one video, image or carousel alongside each playable (Segwise). So the question in the title has a slightly misleading shape. Almost every UA mix needs video. The real question is how much of it should be playable.
What each format does
Video runs everywhere the audience is. It's the native format of Meta, TikTok and YouTube, it goes into the video slots of every ad network, and it's cheap enough to make in the numbers a weekly test needs. Video made up 80.9% of mobile game ad creatives in the first three quarters of 2025, images about 14% and playables 5% (ContentGrip, from Mobvista's XMP and Insightrackr). Most openings a team tests each week are video, because a new opening is a new first three seconds, not a new piece of software.
A playable hands the player the core loop for a few seconds, so the people who install have already played. The small share hides its weight: Liftoff's Creative Index found playables delivering 8 to 16 times the impression-to-install rate of other formats (Segwise). Playables cost more to make and change, and they have to fit each network's rules. AppLovin, Unity Ads, Google Ads, Mintegral, Liftoff, Moloco, Meta and TikTok all accept playable uploads, mostly within 5 MB, and each network calls the store in its own way.
Generative AI is still early for both formats, according to Liftoff, even though 56% of the top-100 grossing games already used it for ad assets in 2025 (Liftoff).
How to decide your mix
Lead with playables if your game's fun is hard to show and easy to feel, a physics puzzle, a satisfying merge, a control scheme that has to be tried, and if playables already carry a large share of your spend on AppLovin or Unity.
Lead with video if your hook is visual or narrative, if you test openings every week, or if most of your spend is on Meta, TikTok and YouTube.
Either way, budget for video around the playables: lead-ins for Meta, companions for TikTok, and the weekly openings and event batches no playable team has time for.
If you're a small studio that hasn't run playables yet, start with video, where the volume and the testing are (testing openings before you commit spend), and add playables when the core loop is the thing your video can't sell.
What Motn makes, and when to keep your playable vendor
Motn makes the video side of the mix: install ads from the studio's own gameplay (why your own footage beats a generated ad), openings to test, offers that change every week, every placement and language, and the lead-in videos Meta needs for a playable. Simple playables for small studios are coming. Motn doesn't claim to match a specialist on playables, and won't until it can show it.
If playables carry a large share of your spend, keep the team or vendor that makes them. Layer and Sett both make playables, and Sett's agents code each one and iterate it as part of a managed service. Use Motn for the video volume around them, so the playable team spends its time on playables. A direct comparison with a tool that does both is in Motn vs Layer.



