What is a Playable Ad?

A playable ad is a user-acquisition format: the user does not watch a video or tap a static — they spend 10–30 seconds in a stripped HTML5 demo of the product, then see an end card with a CTA to the store or a signup. It is try-before-install creative, not a buying channel and not a billing model.

Playables almost always run in mobile in-app (interstitial, rewarded slot, sometimes native), less often in social and in Yandex Direct. The in-app article covers supply inside an app (SDK, CPI, mediation). A playable is one creative type on that supply. Mixing them is like mixing “banner” with “the display network.”

What a playable is not

Rewarded / offerwall traffic is a user completing an action for a reward (coins, premium, a gift card). Rewarded video can sit in the same SDK slot as a playable; the mechanic is different: there they trade attention for a bonus, here they try the advertised game’s loop. An offerwall (Tapjoy, AdGate) is a list of offers, not an interactive demo.

Undisclosed incent in the IO is incentivized traffic, not a playable. A clean playable does not promise “100 coins for installing a competitor”; it hands over a slice of its own game. If an offerwall still follows the demo, that is a mixed source — cut it by sub and by D1, not by format name.

How the creative is built

The usual file is HTML5 (Zip / MRAID) with a tight weight cap (often 2–5 MB on networks, tighter on Meta/Direct). Inside: one or two mechanics from the game, a 5–15s tutorial, a win state, an end card with a button. The player may quit early — the end card must still fire on a timer. iOS and Android need separate builds or a responsive layout; audio is off by default.

Production costs more than static or a simple video: you need a developer or a playable maker, per-network builds, QA on low-end devices. One mechanic hook then yields dozens of variants (difficulty, character, first level) without a new shoot. In a split, cut on play rate (who started), interaction rate, store CTR, and post-install D1 — not CPI in the cabinet alone.

Where it is bought: Unity, Meta, Direct

Unity Ads / Unity Grow. The historical home of playables: casual and mid-core games, interstitial and rewarded. The cabinet takes HTML5 playables; it competes with ironSource LevelPlay and AppLovin AppDiscovery on the same gaming supply. For game offers this is the first format test.
Meta. Playable is a distinct HTML5 format on app campaigns (not Instant Experience / Canvas, which is a post-click fullscreen landing). The user plays in-feed or full-screen in the ad; the CTA goes to the store. Supply is social, not an in-game SDK. It fits when a game or utility is sold to a broad audience, not only to players inside other titles. Weight caps and moderation are stricter than on gaming networks.
Yandex Direct. Playables in display / YAN placements and in CIS app campaigns: HTML5 creative, separate file and store-link rules. This is not “another Unity” — it is a CIS surface under Yandex policy. For RU/KZ app UA it is a scarce but usable slot when static ads stall on CR.

The same file almost never uploads as-is into three cabinets: specs, MRAID wrappers, and click / impression macros differ. Mintegral, AdMob, and inMobi take playables in their SDKs; for the buyer the rules match Unity — weight, end card, store URL.

Metrics and install quality

Buying is usually CPI or oCPI / ROAS; the slot itself may bill CPM. Playables often cost more CPI than static and deliver better D1 / tutorial complete: the person already touched the loop. That is not automatic — a laggy demo, a paywall inside the playable, or the wrong genre converts to install worse than video. Read the chain: play rate → store CTR → install → D1/D7 and LTV, not one network number.

Fraud is the rest of in-app fraud: click injection, device farms, install-without-open. Reconcile with the MMP against install fraud. On iOS SKAN will not tell you which playable won unless conversion value is mapped to creative — you cut by GEO and ad set, not by file.

When the format fits

Strong fit: hyper-casual and mid-core, utilities with one obvious mechanic (cleaner, VPN timer, photo effect), subscriptions you can “touch” in a trial. Weak fit: nutra and finance with no app — there is nothing to play; long-form VSLs belong in ordinary video. In the mobile apps vertical a playable is a UA tool, not a landing replacement: after the store the app still has to match the demo, or D1 drops and the network cuts quality score.