End screens look like a polish detail — those 20-second cards that float at the end of a video. They're actually one of the highest-ROI revenue levers on the platform. A well-designed end screen lifts not just your next video's views but the original video's RPM, because YouTube rewards session-positive uploads with higher per-impression bids.
How session time becomes ad revenue
YouTube's auction prices impressions partly on the predicted "session lift" of the video they're attached to. A video that consistently sends viewers deeper into the platform — more videos watched, longer total session — earns a quiet RPM premium. The premium is invisible in Studio but typically 10–25% over baseline for similar content.
End screens are the most direct lever on session lift. A high-CTR end screen sends 8–15% of viewers into a second video on your channel. Without one, that share drops to 2–4% (the YouTube auto-suggest baseline).
The four end-screen elements that actually convert
- "Best for viewer" video card: Algorithmic pick. Use it. Outperforms manual picks ~60% of the time because YouTube knows what the specific viewer wants next.
- "Recent upload" card: Use only on videos under 30 days old. After that, recency is hurting you — point to evergreen instead.
- Manual video card: Use for a sequel/series pointer. Specificity matters more than freshness here.
- Subscribe element: Always include. CTR is low (0.5–1%) but compounding — and subscribers have 2–4× the lifetime ad revenue of non-subs because of higher return-visit frequency.
The 20-second rule
End screens require at least 20 seconds of video time to render. Most creators end videos cold, lose those 20 seconds, then bolt end screens onto silence. AVD drops, the algorithm reads the dead tail as low quality, and the RPM premium disappears.
The fix: design the end-screen segment as part of the video. A recap, a teaser for the next video, a question to camera, soft outro music under voiceover. Anything that holds attention through the 20 seconds the end screen needs.
Why end screens beat cards (and outros beat both)
Cards (the little "i" pop-up) have CTRs around 0.2%. End screens have CTRs around 5–10%. Spoken outros ("if you liked this, the next video is right here") combined with an end screen card can hit 15–25% CTR on the same card.
The compounding effect: every 1 percentage point of end-screen CTR lift translates to roughly 1.5–2% lift in next-video views, which means another round of end-screen clicks at the new video. Two-tier compounding is the difference between "viral" videos that spike-and-die and videos that pull the rest of the channel up with them.
Pitfalls to avoid
- Overlapping with mid-roll-ad placement. Don't put a manual mid-roll inside the last 30 seconds — it cannibalizes end-screen CTR and the ad doesn't pay enough to make up the loss.
- Pointing to videos with lower AVD than the current one. You're trading good session signal for bad. Always end-screen "upward" to higher-retention content.
- Generic "watch more" voiceovers. "Watch this next" beats "watch more videos on my channel" by 3–4× in CTR. Specificity wins.
How to measure the lift
In Studio, pull "Card click rate" and "End screen element click rate" for your last 30 videos. Sort by RPM. The correlation is usually visible without statistics — top-RPM videos almost always have above-median end-screen CTR. The exceptions are usually videos that got an external traffic spike (skewing the denominator) and aren't representative.
Then run the CTR Revenue Calculator to model what a 5-point CTR lift would do across your top 20 videos. The number is usually larger than any thumbnail-optimization project you could run for the same effort.