End-card design: the last two seconds decide the install

Ask most production teams how long they spent on the interactive middle of a playable and how long they spent on the end card, and the ratio will usually be embarrassing. The middle gets storyboarded and prototyped, then iterated across several builds; the end card gets added in the last week, sometimes by whoever is free, using whatever asset the network's template happens to support. That's backwards. The end card is the only screen in the entire playable where what the player does after tapping is install the game rather than keep playing an ad.

The end-card checklist

Here is a checklist worth running against every end card before it ships, not just the ones going to a new network.

First, timing. The transition from gameplay to end card should happen within a second of the win or loss state resolving, not after a lingering animation the player has already stopped watching. Every extra second between the moment a player decides they've finished and the moment they see the install button is a second in which they can close the ad instead.

Second, the call to action has to be one unambiguous button. Not a button competing with a rating prompt or a second interactive element; multiple tappable elements at the end of a playable measurably split intent, and the network's own close button is already competing for that same thumb.

Third, the message on the end card should match what the player just did rather than restate the game's pitch from the opening screen. A player who just solved a puzzle wants to see more puzzles, not a generic "download now" over unrelated footage, and reusing opener messaging here wastes the one moment the player has demonstrated real engagement.

Fourth, test the end card in isolation from the rest of the playable. A strong end card can sometimes rescue a weak middle, and a confusing one can undo a strong middle; if the team only ever evaluates end cards bundled with the full playable's performance, it never learns which part drives the outcome. Run end-card variants against a fixed middle when the budget allows.

Fifth, check load behaviour on the exact devices your media mix reaches, not just flagship test units. An end card that hangs for even half a second on a mid-tier Android device after a smooth gameplay sequence reads as a broken ad. Players close broken ads.

Sixth, localise the end card's text and any voiceover with the same care as the gameplay instructions, not as an afterthought translated by whoever is available, because a grammatically awkward call to action undercuts trust at the exact moment a player is deciding whether to trust the install prompt.

Seventh, watch for network-specific end-card behaviour rather than assuming one build works everywhere. Some networks wrap a playable's end card with their own close button and rating prompt, or a store badge, layered on top of whatever the production team designed, so an end card tested on one network's SDK can look completely different once it ships through another's template. Confirm what each network adds before treating a single build as final across the whole media mix. Don't wait for a campaign's conversion rate to come back low on one placement to find out.

Eighth, keep a record of which end-card variants moved conversion, kept separate from the list of ones the team simply liked, and keep it short. It's easy for an end card to become the place where a producer's personal preference for a colour or a line of copy gets baked in without a test behind it, because by the time production reaches the end card everyone's tired and the deadline is close. A short log of what the team tried and what happened stops the following production cycle from repeating a choice that already failed once.

Why this is worth two hours per playable

Put together, this checklist is closer to two hours of dedicated review per playable than to a redesign. That's a small cost against what the end card is responsible for: converting a player who has already given the ad thirty seconds of attention into an install, rather than losing them in the last two.

None of this requires new tooling or a bigger build. It requires treating the end card as its own deliverable with its own QA pass, on the same schedule as the interactive sequence that earns the player's attention in the first place.

Related archive reading

These articles provide related context and remain subject to their stated review status.

• End cards and CTAs: the neglected frame
• The Playable Ad QA Checklist to Run Before Submission