YouTube End Screens: Layout, Timing, and CTA Examples

Most creators treat the end screen as an afterthought — the video ends, they drag a template into the editor, and whatever was on screen gets covered by two thumbnails and a subscribe bubble. Then they wonder why nobody clicks.

The end screen is the last piece of persuasion you get. A viewer who made it to the final seconds is your warmest audience: they finished, which means the video worked on them. Sending that person to the wrong next video, or to nothing at all, is leaving the most qualified click on the table. This guide covers the mechanics YouTube actually enforces, the layout patterns that fit different video types, how to write the verbal CTA that makes the end screen work, the mistakes that quietly destroy click-through, and where to find the numbers in Studio.

---

The rules: what YouTube actually gives you

Before designing anything, know the constraints — they're published by YouTube and they shape every decision below:

Two practical consequences. First, plan the last 20 seconds in the edit, not after the upload — YouTube's own guidance is to leave enough space and time at the end while you're cutting. Second, because viewers watch on phones, tablets, and TVs, keep elements inside the middle of the frame: text, faces, and critical visuals should avoid the corners where overlays land, and nothing important should sit under where a subscribe button goes.

---

The element options: what each one is for

You get these building blocks in YouTube Studio's editor (Content → your video → Editor → End screen):

The video element has a hidden choice: "recent upload," "best for viewer," or a specific video. "Best for viewer" sounds smart and often underperforms a deliberate pick — YouTube optimizes for its own relevance guess, not for the video that continues *your* narrative. For a tutorial series, choose the specific next episode. Save "best for viewer" for channels where any video is a reasonable next watch.

How many elements? Two or three is the practical ceiling for most layouts. Four elements technically allowed doesn't mean four elements worth clicking — every additional target splits attention, and a viewer scanning four thumbnails often closes the video instead of choosing. One strong next video beats a menu.

---

Layout patterns that work

The layout is a bet on what a satisfied viewer should do next. Match the pattern to the video's job:

Whatever pattern you choose, three placement rules apply: keep elements off faces and captions (viewers don't click what they can't read), leave margin from the edges so nothing clips on different screen sizes, and preview before publishing — the editor's play button shows exactly what a viewer sees.

---

Script the verbal CTA into the video

An end screen dropped on a silent outro gets silent results. The elements need a spoken runway, and the script should sound like the last line of the video — not an ad break.

The structure that works:

1. Deliver the final payoff. The last useful sentence lands first. Value never gets postponed for the ask.
2. Bridge with "if/then" logic. Tie the recommendation to what they just got: *"If this fixed your audio, the next step is sound treatment — that video's right here."*
3. Make the ask while value is still on screen. Start talking about the next video 10–15 seconds before the elements appear, so the click moment has context rather than dead air.
4. End on direction, not goodbye. The final spoken words should point at an action ("start with that one"), not at a farewell.

Three CTA examples you can adapt:

• Tutorial: "That's the whole fix — restart, and your audio should be clean. I've put the companion video on mic placement right here; watch that next if you're still hearing room noise. And if you want the rest of this setup series, hit subscribe — I break down one studio problem every week."
• Listicle / review: "If you only try one of these, make it the second pick — cheapest and the one I actually use. The full review of it is over here, and the playlist below has the rest of the series."
• Story / essay: "Which is why the ending mattered more than the middle. I made a follow-up that goes into what happened after — it's here. Thanks for watching this one to the end."

Notice none of them say "don't forget to like and subscribe" as an opener. The subscribe ask rides *second*, after the next-video recommendation, because continuing to watch is the stronger commitment signal — and the stronger gift to your channel.

For the CTA to sound scripted rather than improvised, write it into your script document as its own section — the same discipline the YouTube script formula for retention applies to hooks. An outro written at 1 a.m. in the edit is an outro that says "uh, yeah, that's it."

---

Common mistakes that kill end screen clicks

Clutter. Four elements, all equally sized, competing with captions and a watermark (cards and watermarks are suppressed anyway — don't design around them). The viewer's eye never lands, so it leaves. Pick the one action that matters most and give it room.

The mismatched recommendation. Your video is about fixing audio; the end screen auto-fills "best for viewer," which surfaces your vlog from March. A satisfied viewer clicks the wrong thing, watches 40 seconds, and leaves — you converted a good session into a bounce. Specify the video when the narrative has a correct next step.

The too-long outro. A 40-second farewell where elements only appear in the last 15 seconds: viewers drop during the speech and never see the click moment. Keep the closing tight — the end screen window should feel like part of the video, not a commercial attached to it.

The dead-air ending. Video content stops, three seconds of music play, then elements fade in. Anyone still watching has nothing to look at except the thumbnails — and silence kills more clicks than bad thumbnails do. Keep talking until the elements arrive.

Designing for desktop only. On a phone, your elegant two-column layout sits on a smaller canvas with fingers hovering over it. Test on your own phone; if a tap target looks cramped, make it bigger or move it.

Never updating the picks. A hardcoded "next video" from eight months ago, still pointing at a topic your current audience doesn't care about. Revisit the end screen on your top-performing videos each quarter — they keep accruing views long after you've forgotten them.

---

How to measure end screen performance in Studio

You don't need invented benchmarks; you need comparisons. YouTube reports end screen performance in YouTube Studio → Analytics, in the expanded report for the video (and across videos via the end screen cards Studio surfaces). Read it qualitatively at first:

• Which element type earns clicks? Studio's reporting lets you compare element types — if your subscribe element consistently gets clicks and your playlist never does, the playlist is asking the wrong question of your audience.
• Which videos earn end screen clicks? Studio highlights top videos by end screen performance. Look at what those have in common: a tighter verbal CTA? A layout with fewer elements? A recommendation that continues the exact topic?
• Clicks relative to the people who saw it. The report frames clicks against end screen impressions — how many viewers actually reached the window and how many acted. A video with modest views but a high share of viewers clicking is packaging a better next step than a video with big views and no clicks.
• Pattern across your last ten uploads. One video's number is noise. If your last ten tutorials all show the same weak element, that's a fixable pattern — usually layout or CTA, not luck.

Act on the pattern: change one variable (element count, placement, or the spoken CTA) for the next batch of videos, then read the report again. The same comparative habit the analytics beginner's guide teaches for retention applies here — you're looking for the difference your change made, not a number to feel good about.

If your end screens are structurally weak — old videos with no end screen at all, mismatched picks, CTAs never scripted — the YouTube Channel Audit checklist walks you through fixing the backlog one video at a time.

---

Pre-publish end screen checklist

• [ ] Last 20 seconds of the edit leave clean space: no critical text or faces where overlays land.
• [ ] Video is longer than 25 seconds (obvious, but Shorts-adjacent edits sometimes get cut short).
• [ ] Two or three elements maximum, each with a clear job.
• [ ] The video element points at a *specific* next watch when the narrative has one.
• [ ] The verbal CTA starts before elements appear and ends on a direction, not a goodbye.
• [ ] Previewed the end screen in the editor at full playback — nothing overlapping captions or faces.
• [ ] Captions/chapters plan doesn't fight the window: chapters that end the video cleanly give the outro room, and the Timestamp Generator keeps your description timestamps tidy so viewers who jump to the end land where you intended.
• [ ] Added end screens to your older top performers, not just the new upload.

The end screen is small real estate with outsized intent — the only moment in a video where every remaining viewer is a candidate for the next click. Spend the last 20 seconds like you mean it.

---