The YouTube upload screen is a form with roughly a dozen decisions on it, and several of them can cap a video's reach before a single person watches it. The made-for-kids declaration, the visibility setting, the chapter timestamps you didn't paste, the end screen you meant to add later — these are silent failures. Nothing errors out. The video just performs worse than it should, and you never see the counterfactual.
This is a checklist for that form: what each field does, which ones actually matter, and how to set channel-level defaults so you're not re-deciding the same eight things on every upload.
The fields that cost you reach
Here's the short version of what breaks when each setting is wrong.
| Setting | What goes wrong if you skip or fumble it | Where it lives |
|---|---|---|
| Audience (made for kids) | Comments, end screens, cards, notifications and playlist saves all disabled; ads become non-personalized | Details |
| Visibility | Video goes live at 4am, or sits unlisted for a week | Visibility |
| Description (first 2–3 lines) | Lost search context and no above-the-fold reason to keep watching | Details |
| Chapters | No chapter markers in the player, no key-moment surfaces in search | Description |
| Custom thumbnail | YouTube picks an auto-frame from your footage — usually a blink | Details |
| End screens and cards | No path to the next video; the session ends with you | Video elements |
| Language and captions | Weaker accessibility and worse discovery outside your main locale | Details / More options |
| Paid promotion and altered-content disclosures | Policy exposure, and the risk is entirely avoidable | More options |
Step zero: what should already be done before you open Studio
The upload screen is a bad place to make decisions. Every field you improvise there is one you'll want to change after publish — and post-publish edits arrive after the first traffic wave, which is the wave that matters. Have these ready before you touch the upload button:
- The exported file at the right spec. Resolution, aspect ratio and codec sorted at export, not in a re-upload. Our YouTube video size reference covers the dimensions and format targets worth exporting to.
- The thumbnail as a finished image file, matching the YouTube thumbnail size spec. Custom thumbnails need a verified account, so check that once and forget it.
- Title and description written in a doc, not typed live. Frameworks for both are in our YouTube titles and descriptions guide, and you can sanity-check length in the YouTube character counter before pasting.
- Chapter timestamps pulled off the edit timeline. Your editor already knows where the sections start; write them down while you're still in the NLE.
- The playlist decided. Not "I'll sort playlists later" — later never comes. If you don't have a playlist structure yet, playlist strategy is worth twenty minutes.
- The publish slot picked, ideally against your own analytics rather than folklore. Our best time to post on YouTube page is a reasonable starting hypothesis if you don't have data yet.
The failure mode this prevents is familiar to anyone who's uploaded for a client: it's 4:50pm Friday, the file is still rendering, the client wants it live today, so you type a title into the box and promise yourself the description gets fixed Monday. The video's first 48 hours — the window YouTube uses to decide who else sees it — run on a placeholder.
The upload screen, field by field
Details
Title. Paste the one you wrote. If it's getting truncated in the box, it's getting truncated in search and on mobile home too.
Description. The first two or three lines carry the most weight: they're what shows above the "more" fold and they're prime real estate for search context. Put links below that, and remember that raw URLs render as clickable links in the description while text does not — the mechanics are covered in making YouTube description links clickable.
Thumbnail. Upload the custom file. If you're batching, thumbnails can be prepared and attached ahead of publish — see how to add and schedule a YouTube thumbnail.
Playlists. Assign the playlist here, at upload. Retrofitting playlists across 80 back-catalogue videos is a miserable afternoon.
Audience: the made-for-kids flag
This is the single most consequential toggle on the page, and it's presented like a formality. Declaring a video as made for kids is a legal requirement where it applies — but when it's set, YouTube disables comments, disables notifications to subscribers, removes end screens and cards, blocks save-to-playlist, and serves non-personalized ads only (which typically means less revenue).
You can set a channel-level default in Studio's advanced settings, but it should match reality — if only some of your content is directed at children, declare per video. And if you flag a video by accident and only notice a week later, you can change it, but the subscriber notification has already not been sent. There's no re-send.
If your video is not directed at children, set the flag accordingly and move on. If it is, plan the video knowing end screens and comments won't exist for it.
Chapters
Chapters come from timestamps in the description, not from a separate field. YouTube's requirements are specific: the first timestamp must be 00:00, you need at least three timestamps in ascending order, and each chapter needs to be at least ten seconds long. Miss any of those and the chapters simply don't render — no warning, no error.
A working example:
00:00 What this covers
01:12 The three settings people skip
04:30 Setting upload defaults
07:45 What to check after publish
Chapters help viewers navigate long content, and the labels give YouTube more textual signal about what's in the video. For anything over about five minutes with distinct sections, they're free value.
There's also an "allow automatic chapters" option — a reasonable fallback for a back catalogue you'll never retrofit. For anything you care about, write them yourself; your section names will beat the generated ones.
More options: the drawer nobody opens
Expanding "Show more" reveals the settings most creators never touch:
- Paid promotion — tick it when a sponsor is involved. Non-negotiable.
- Altered or synthetic content — disclose realistic AI-generated or meaningfully altered footage. This is a standing requirement, not a new one, and the cost of ticking it honestly is near zero.
- Tags — a minor signal at best. Worth a minute for common misspellings of your brand or topic, not twenty.
- Language and captions — set the video language and upload a caption file if you have one. Auto-captions are rough on jargon and accents.
- Recording date and location — useful for event, travel and local-business content; irrelevant otherwise.
- Category and comment settings — held-for-review beats wide-open on a video you won't be around to moderate.
- License and embedding — leave standard unless you have a reason.
- "Publish to subscriptions feed and notify subscribers" — on by default. Turn it off for re-uploads and housekeeping videos you don't want pushed to everyone.
Video elements
End screens and cards are the difference between a view and a session, and they're the elements most likely to get skipped — they require the video to already be uploaded, so they land at the exact moment you're ready to be finished.
Add them anyway. Our end screens and cards guide covers placement that chains views without wrecking your retention curve. Two constraints worth knowing at upload time: very short videos can't carry an end screen, and Shorts support neither — so for Shorts, the next-video path lives in the description, a pinned comment, or the content itself.
Checks and visibility
The copyright check runs automatically. Wait for it before scheduling — discovering a claim after the video has gone live in front of your notification wave is a bad way to spend an evening. Then the last decision:
| Visibility | Use it for |
|---|---|
| Private | Client review, internal approval, your own QA pass |
| Unlisted | Embeds, shareable links, anything not meant for the feed |
| Public | Immediate publish |
| Scheduled | A specific publish time, with the notification going out then |
| Premiere | A live countdown and chat moment around the drop |
Premieres are worth considering for launches and anything where an audience showing up together adds something — the setup is covered in scheduling a YouTube premiere.
One processing note: high-resolution processing can lag behind the standard-definition version. If you publish the instant the upload bar fills, your earliest viewers — the ones whose watch behaviour weighs most — may get a soft-looking video. Scheduling an hour or two out avoids this entirely.
Upload defaults: stop re-deciding the same eight fields
In YouTube Studio, Settings → Upload defaults pre-fills the fields that rarely change: description boilerplate, tags, category, language, comment settings, license and default visibility. Set them once and every upload starts most of the way done.
Worth putting there: a description template with your standard footer, your default category and video language, your comment moderation setting, and — importantly — default visibility set to private rather than public. A private default makes an accidental publish impossible; you promote to scheduled or public deliberately.
Leave the audience declaration out of your defaults unless every video you make genuinely has the same answer. That's the one field that deserves a fresh decision each time.
The ceilings nobody warns you about
Unverified accounts are capped at short video lengths and can't upload custom thumbnails — verify once, ideally before you need to.
YouTube also limits how many videos a channel can upload in a 24-hour window. The number isn't published and varies with channel age, verification status and history, which matters most during a back-catalogue migration or a Shorts sprint. Spread uploads across days rather than dumping thirty files on a Tuesday and discovering the ceiling at file nineteen.
Where scheduling fits
A scheduling layer helps here for one reason: sequencing. When a video is scheduled rather than published on completion, the metadata has to exist before the publish slot arrives. Title, description, thumbnail and chapters get written during planning, when you have time to write them well, instead of being patched in after a rushed upload.
That's the honest case for scheduling YouTube through SocialKit: compose once, customise the caption per platform if the clip is also going elsewhere, and see it in a visual calendar next to the rest of the week. Auto-publish handles the drop; post analytics tell you how it did. YouTube (including Shorts) is one of the eleven platforms every plan includes — from €29/month Solo, or €17.40/month billed annually as of July 2026, with a 7-day trial.
What it doesn't do: end screens, cards and copyright checks are Studio-native and always will be, and SocialKit has no comment-moderation queue or social inbox, so replies stay in Studio too. The split that works in practice is scheduling and metadata in the calendar, elements and moderation in Studio — the full flow is in how to schedule YouTube videos.
Your pre-publish checklist
Run this in order. It takes about four minutes once it's habit.
- File exported to spec, thumbnail exported as a separate image.
- Title and description pasted from your doc, not typed live.
- Chapters pasted, starting at
00:00, three or more, ten seconds minimum each. - Thumbnail uploaded and confirmed — not the auto-frame.
- Audience flag answered deliberately for this specific video.
- Playlist assigned now, not later.
- More options opened: paid promotion, altered content, language, captions, category, comments.
- End screen and cards placed, or consciously skipped because it's a Short.
- Copyright check finished clean.
- Visibility set to a scheduled slot, at least an hour out so processing completes.
After it publishes, one last pass: watch the first thirty seconds on your phone, confirm the thumbnail renders, confirm chapters appear in the scrubber, and check the description links work. Then leave it alone — the first 48 hours of YouTube analytics tell you more than any edit you'd make in hour three.