End cards are the last 15 to 20 seconds of a video where platforms like YouTube let creators overlay clickable elements: subscribe buttons, recommended videos, links to external sites. They're almost universally treated as free real estate. Put your branding up, drop in a link, and let the platform do the rest. The problem is that every element placed in those final seconds sends a visual signal that the video is finished. Viewers notice. They leave.
Completion rate is one of the metrics platforms weight most heavily when deciding which videos to recommend. A video that loses 30 percent of its audience in the last 20 seconds sends a clear signal that something in that window caused the exit. Most creators assume viewers stopped because the content ran long. The end card is rarely examined, and it's often the cause.
What end cards actually look like to a viewer
A viewer watching a 4-minute explainer doesn't know the runtime before they start. They judge where they are by visual and structural cues: the pacing of the edit, the rhythm of the narration, the music. When a grid of clickable thumbnails slides in from the corner at the 3:40 mark, that grid communicates "this is over" as clearly as rolling credits. The brain registers it as a boundary, and a viewer who was still engaged disengages.
This is closely related to a principle West Melbourne Studios has explored when looking at how attention metrics are replacing view counts in digital video: the final seconds of a video disproportionately shape whether a viewer is counted as having truly watched. Platforms that track heat maps of engagement consistently show a drop at the moment the end card appears, not at the moment the video ends.
The visual intrusion problem is compounded on mobile. A thumbnail panel that looks reasonable on a desktop browser covers a significant portion of a phone screen. If the video's closing argument or final visual is obscured by a subscriber button, the viewer's last impression isn't the brand's intended frame. It's a white circle with a bell icon.
The timing problem most creators ignore
YouTube requires end cards to appear in the final 20 seconds of a video. Most creators set them at exactly 20 seconds because that's the default. It's rarely the right choice.
A 90-second video with a 20-second end card has roughly 22 percent of its runtime devoted to promotional overlay. That's a different proposition to a 10-minute documentary where 20 seconds represents 3.3 percent. The platform rule sets a maximum, not a mandate. Setting end cards to appear at the 10-second mark instead of 20 seconds can recover significant viewing time, and on shorter videos it's often the difference between a completion and a bail.
The audio matters here too. If a video ends with a narrated sentence that finishes at the 15-second mark and then holds on a branded frame in silence, the end card appearing at second 12 is less disruptive than it would be mid-sentence. Structure the closing sequence with the end card placement in mind, not as an afterthought in post.
How many elements is too many
YouTube allows up to 4 end card elements per video. Most creators use all 4, partly because the option is there and partly because the platform's own tutorial encourages it. Four clickable panels competing for attention in the final seconds of a video is visual noise, and visual noise drives clicks away from everything, including the video itself.
The strongest end card strategies use 2 elements at most, positioned away from the subject's eyeline if there's a presenter on screen, and introduced only after the last word of narration has landed. The design principle is the same as any good call to action: fewer options produce clearer decisions. This connects directly to the guidance West Melbourne Studios has outlined on how split testing video length affects audience retention. The same logic that applies to runtime applies to the end card window. Test fewer variables, not more.
The brand perception cost nobody measures
Beyond completion rates, poorly placed end cards carry a subtler cost. A viewer who was genuinely moved by the final 30 seconds of a brand video, and then had that moment interrupted by a hovering subscribe icon, doesn't leave with the brand's intended emotional impression. They leave with an experience that felt slightly off. They might not articulate why. They just don't share it.
West Melbourne Studios works with clients specifically on the architecture of closing sequences in brand and corporate video, because the final moment of a video is the one a viewer carries away. How video buffering ruins brand perception is a documented problem; end card disruption works on the same mechanism. The technical or structural element interrupts the emotional arc and leaves the viewer slightly dissatisfied without knowing exactly why.
There's no platform rule requiring end cards at all. For brand films, documentary content, or high-production corporate video, removing end cards entirely and letting the final frame breathe is often the better commercial decision. The recommended video that appears after a clean ending is served by the platform anyway. The end card just gets in the way earlier.
What good end card design actually looks like
Good end card practice starts before the edit. The closing sequence is scripted with the overlay in mind. If the final image is a wide shot with empty space in the upper right, that's where the element lands, not over a face or a logo.
Timing is set to the latest defensible point, which for most videos is between 10 and 14 seconds before the end. One or two elements only, with the primary call to action given clear visual priority. If the video has a single goal, the end card reflects one goal.
For YouTube creators and brand channels publishing regularly, A/B testing end card placement is straightforward using YouTube Studio's built-in analytics. Watch-time graphs show the exact second viewers are leaving, and overlaying that against end card trigger points reveals whether the card is contributing to drop-off. It's one of the few areas of video optimisation where the data is both free and immediate.
The end card is a tool, not a right. Used with the same care as any other element in a video, it can add genuine value. Used as an afterthought, it quietly undoes the work of every minute that came before it.