YouTube Pre-Publish Checklist (2026)

Twenty-seven numbered steps, plus one check before you start. Each row states what passing looks like. Where a threshold is measured, the source is linked; where no published threshold exists, the row states the check instead of inventing a number.

Each phase below explains how to run its steps and what to listen or look for.

This is the highest-leverage phase. A problem caught here costs minutes to fix. The same problem caught after filming costs a reshoot, or ships in a video that underperforms.

Step 0: Check the topic against your comments. Before you check how the script is written, check whether anyone asked for it. The Comment Miner groups your public comments into the requests and questions viewers keep raising. A well-written script on a topic nobody asked about is the most expensive kind of mistake to catch late.

Step 1: Run a script analysis. Paste the complete draft into a script analyzer such as Prepublish. Read the overall score first, then check whether the hook is your strongest section rather than your weakest. The bands are published in the product: 80 and above is strong, 60 to 79 is good, 40 to 59 needs work, and below 40 is weak (prepublish-fe/lib/score-bands.ts). Then read the script-hold map for sections that stand out against the rest of the same draft. Treat that map as a comparison between drafts of one script, not as a claim about published viewers. If the draft sits in the needs-work band, rewrite before you film.

This step has a deeper version. The check your YouTube script before recording guide runs the same check as five separate tests with the evidence behind each. Use that guide when one draft needs diagnosis. Use this checklist when you want the pass that covers the whole production.

Step 2: Validate the hook. Read your opening aloud. The check is not a stopwatch, it is a question: does the first sentence carry the claim? Our 349-video hook study found the timing of the first concrete reason to keep watching was nearly identical between hits and flops (median word 14 against word 15), so early delivery on its own separates nothing. What did separate them was whether the opening engaged the title's promise: 87 percent of overperformer openings against 78 percent of flops.

Fail indicators:

• Your first sentence is a greeting ("Hey guys, welcome back to the channel")
• The opening is context-setting ("So I have been thinking a lot about this topic lately and I wanted to share my thoughts")
• The opening explains what the video is about instead of why it matters ("In this video, I am going to walk you through five tips for...")

Pass indicators:

• You open with a specific, unexpected claim ("I have posted 400 videos. Only 3 broke 100K views. They all had one thing in common.")
• You open with a result or transformation ("This one camera setting took my videos from 200 views to 20,000.")
• You open by demonstrating the problem the video solves ("Watch this clip. See how the viewer drops off at exactly 47 seconds? That is not random. There is a specific reason.")

For more hook patterns and examples, see our hook examples guide and first 30 seconds guide.

Step 3: Check title-script alignment. Read your title. Then read your opening. Does the script begin delivering the title's promise before any credentials, context or setup?

If your title is "5 Camera Settings Most Beginners Get Wrong," your first setting must appear straight away, not after a preamble about why camera settings matter.

If your title is "I Tried X for 30 Days: Here's What Happened," the results, or at least the first surprising result, must appear in the opening. The 30-day story setup comes after.

The benchmark: a viewer who clicked because of your title should still be confident the video will deliver by the end of the opening. If the viewer would be wondering "when does this actually start?", the script fails this step. The 349-video hook study measured this signal inside the first 45 seconds of the opening, which is the window to check.

Step 4: Audit your transitions. Read through every point where your script moves from one section to the next. For each transition, check: does it create forward momentum (adversative) or signal more of the same (additive)?

Forward momentum transitions (use these):

• "But here's where it gets interesting..."
• "The problem is, most people do this wrong..."
• "Now, you might think [X], but actually..."
• "This works great, until you try it with [Y]..."

Flat or backward transitions (replace these):

• "Moving on to the next point..."
• "Another thing you should know..."
• "Also..."
• "On top of that..."
• "The next tip is..."

Benchmark: most of your transitions should be adversative or forward-leaning. A run of additive transitions makes the middle feel like a list instead of an argument. Our script structure guide goes deeper on transitions and pacing.

Step 5: Check re-engagement placement. Mark the middle of each long stretch of your script. Each one should contain a deliberate re-engagement moment: a sentence or short paragraph that re-hooks the viewer. Three types work:

• Stakes escalation: "But this next part is the one that actually costs people money." (Raises the stakes of what's coming)
• Open loop: "I'll show you the exact settings in a minute, but first you need to understand why they work." (Creates a question the viewer needs to stay to answer)
• Unexpected pivot: "Now, everything I just said? It's wrong for one specific situation." (Subverts expectations and re-engages curiosity)

Benchmark: no long stretch of the body should pass without one of these. A section that runs on without a fresh reason to keep watching is where the middle of the video loses people.

Step 6: Verify the payoff. Read your script's closing section. Does it:

• Actually deliver the conclusion or answer promised by the hook? (Not a summary of what was covered, but the actual payoff)
• Include one unexpected piece of value (a bonus tip, a reframe, a "one more thing") that rewards viewers who stayed to the end?
• End on a forward-leaning note that makes the viewer want to watch another video?

Benchmark: the closing section should contain at least one piece of information that is not predictable from the rest of the video. If a viewer could stop early and have everything, the ending is too weak.

Step 7: Time your script. Read your full script aloud at your natural speaking pace and time it. Across 349 measured videos, the middle half of creators spoke between 160 and 201 words per minute, with a median of 181 (prepublish-fe/lib/seo/wpm-data.ts, and the same figures appear on words to minutes). Where a number comes from a study it is linked, and the rest is a working convention. For what finished videos hold, the corpus median for a 12-minute video is 2,208 words, with the middle half between 1,920 and 2,436 (word count targets, the statistics page).

Compare the estimate to your target. If the draft runs well over, cut now instead of planning to "edit it tight." If it runs well under, a section is missing or thin.

Benchmark: the script's estimated spoken length should land inside the 15 percent buffer the planning table uses. Outside that range, restructure before recording.

Step 8: Record your hook first and play it back. Film your opening. Stop. Play it back on your phone, not your camera's viewfinder, because your phone simulates how viewers will actually watch. Ask yourself: "If I saw this in my feed with no context, would I keep watching?"

If the answer is not an immediate, honest yes, film it again. The opening deserves more takes than anything else in the video.

Benchmark: your hook take should make you want to keep watching even though you already know what the video contains. If you are bored by your own opening, your audience will be gone.

Step 9: Check audio before full recording (the earbuds test). Record a short take of yourself speaking in your filming setup. Play it back through wired earbuds, not speakers, because earbuds reveal problems speakers hide. Listen for:

• Room echo or reverb (sounds hollow or distant)
• Background noise (AC hum, traffic, computer fans)
• Mic pops on P and B sounds
• Volume: your voice should be clearly audible at a normal listening level
• Sibilance: harsh S sounds that cut through

Benchmark: your audio should sound clean enough that a viewer could listen to the video as a podcast without being distracted by audio quality. If it does not pass that test, fix your recording setup before proceeding.

Step 10: Plan your energy management. If your recording session will be a long one, plan specific break points. Energy falls during sustained recording. Viewers can feel low energy even when the script words are perfect; it shows up as flat delivery, slower pace, and less vocal variety.

Mark break points in your script where you will stop, stand up, walk around, then resume. These should align with major section transitions.

Benchmark: the energy level in your final section should be at least as high as your opening. If your endings consistently feel flat, energy management during filming is the fix.

Step 11: Film pattern interrupts. If your script calls for demonstrations, screen recordings, location changes, or physical props, make a shot list and check each one off during filming. It is extremely easy to default to talking head for the entire video when you are in the flow.

Visual variety is not optional for retention. Viewers need a visual change often enough that the edit has something to cut to. Our pattern interrupt cadence model puts the sweet spot at one interrupt every 60 to 75 seconds, tightening through the first minute.

Benchmark: for every stretch of finished video, you should have at least one alternative shot or visual element available. If you only have talking head footage, your editing options are limited to jump cuts and zooms, which work for some formats but get monotonous quickly.

Step 12: Watch at normal speed (the drift test). Before making a single edit, watch your raw footage at normal speed from start to finish. Not faster. Not skipping around. Every moment where your attention drifts (where you reach for your phone, where you think about something else, where you want to skip ahead), mark that timestamp.

Those drift points are where your audience will leave. They are the most important data points in your entire editing process.

Benchmark: if your own attention drifts repeatedly through the raw footage, you have structural problems that editing alone cannot fix. You may need to restructure sections or cut entire segments.

Step 13: Tighten the opening with aggressive editing. Your opening should be the most aggressively edited section of your entire video. Cut every unnecessary word, pause, breath, filler, and hesitation. Remove:

• "So, basically..."
• "Um" and "uh"
• Throat clearing
• Any sentence that restates the hook in a weaker way
• Any context that the viewer does not need yet

The pacing of your opening sets the viewer's expectation for the rest of the video. If the opening feels fast and tight, viewers will be patient during slightly slower sections later. If the opening feels slow, viewers will assume the whole video is slow.

Benchmark: your opening should contain no dead air and no filler words. Use jump cuts aggressively here even if you use them sparingly elsewhere. For a deep dive on optimizing your opening, see our first 30 seconds guide.

Step 14: Visual variety audit (the timeline scrub). Scrub through your editing timeline and look at the visual track, not the audio. Note the stretch between distinct visual changes (different shot, b-roll, text overlay, zoom, screen recording, and so on).

Benchmark: no single visual should persist long enough to stop registering as a shot. Even a subtle zoom on the same talking head shot counts as a change; it is enough to re-engage the viewer's visual attention. The cadence numbers we publish live in the pattern interrupt guide, where the sweet spot is one interrupt every 60 to 75 seconds.

Step 15: Audio consistency check. Play through the full video and listen specifically for:

• Volume jumps between sections, especially if you recorded on different days or in different locations
• Background noise changes (fans, outdoor sounds appearing and disappearing between cuts)
• Music volume relative to voice (music should sit under the voice, never over it)
• Overall loudness consistency from the first section to the last

Benchmark: a viewer should not need to adjust their volume at any point during your video. If they do, your audio is inconsistent.

Step 16: End screen window. YouTube places end screens in the last 5 to 20 seconds of a video, and the video has to be at least 25 seconds long to have one (YouTube Help). Verify:

• Your video has a clean outro section that accommodates end screen elements
• No important visuals or text appear in the areas where end screen cards will overlay (the right side and center of the frame)
• Your verbal CTA ("check out this video next") aligns with when the end screen appears
• The end screen elements run for most of the window rather than a few seconds of it

Benchmark: end screen elements should appear while you are still delivering value or your verbal CTA, not during silence or a static outro card. An end screen over dead air gives the viewer nothing to click.

Step 17: Write your title last (the strongest-angle test). Your title should reflect the actual video you made, not the video you planned to make. These often diverge during production. The angle that felt strongest during scripting may not be the angle that came alive during filming.

Watch your finished video and identify: what is the single most compelling moment or insight? That should be your title's focus.

Title formula checklist:

• Short enough to survive truncation in search results and the mobile feed
• Contains your primary keyword early in the title
• Uses a number if applicable (a specific number is easier to click than a vague plural)
• Creates a specific curiosity gap or promises specific value
• Does NOT over-promise what the video delivers (clickbait hurts retention, which hurts the algorithm)
• You would click on this title from a creator you do not follow

Benchmark: read your title and imagine seeing it from a stranger's channel. Would you click? Would you feel the title delivered after watching? Both answers must be yes.

Step 18: Thumbnail-title complementarity. Your thumbnail and title work as a single unit. They should complement each other, not repeat each other.

Bad pair: Title "5 Biggest YouTube Mistakes" + Thumbnail text "5 MISTAKES" Good pair: Title "5 Biggest YouTube Mistakes" + Thumbnail showing a creator's shocked face with an analytics graph going down

The thumbnail adds emotion, visual context, or curiosity that the title cannot convey in text. The title adds specificity that the thumbnail cannot convey visually.

Benchmark: cover your title and look at just the thumbnail. Then cover the thumbnail and read just the title. Each should make sense independently but neither should give the full picture alone. Together, they should be more compelling than either one separately.

Step 19: In the description, front-load the lines viewers see. Only the opening lines of your description appear in search results and above the fold on the watch page. These lines must:

• Contain your primary keyword naturally
• Give a compelling reason to watch (not just describe the video)
• NOT start with "In this video, I..." (this wastes the most valuable real estate in your description)

The rest of the description should include:

• Timestamps for each major section (these also appear as chapters in the player)
• Links to related videos mentioned in the content
• Your primary keyword plus a few close variations woven naturally into the prose
• A call-to-action

Benchmark: your first description line should be strong enough to work as a post that makes people want to watch. If it reads like a table of contents, rewrite it.

Step 20: Tag the video. Tags influence YouTube's initial content classification. Use:

• Your exact primary keyword as the first tag
• Close variations of your primary keyword
• Broader category tags
• Your channel name

Benchmark: every tag should be a phrase someone would actually search for. If you would not type it into YouTube's search bar, it is not a useful tag.

Step 21: Category and language settings. Select the YouTube category that most closely matches your content. This affects which audiences YouTube initially tests your video with. Common mapping:

• Tech reviews -> Science & Technology
• Gaming -> Gaming
• Tutorial/educational -> Education or Howto & Style
• Commentary/essay -> Entertainment or People & Blogs
• Finance -> Education

Also verify your language setting matches your video's spoken language. Incorrect language settings can cause your video to be served to audiences who cannot understand it, which hurts retention.

Step 22: Schedule ahead of the audience window. In YouTube Studio, open Analytics and then the Audience tab to see when your subscribers are most often on YouTube. Schedule your upload ahead of that window, with enough time for the video to finish processing and for YouTube to start distributing it to your subscribers as they come online.

If you do not have enough data for audience insights, pick a time you can be present to answer comments during the first hours after publishing, and keep it consistent.

Benchmark: your video should be fully processed and available before your audience's peak activity window, not during it. Uploading during the peak means your early viewers meet a processing or low-quality version.

Step 23: Community post (the pre-launch prime). Write a community tab post to publish shortly before your video. This primes your subscriber base. Include:

• A still from the video (not the thumbnail; a different, curiosity-generating frame)
• A question related to the video's topic that invites comments
• Do NOT include the video link yet; just generate anticipation

Then, when the video publishes, add a second community post with the link and a brief "it is live" message.

Benchmark: your pre-launch post should earn replies before the video goes live. If it does not, the question was not specific enough to answer.

Step 24: First comment (the engagement seed). Before publishing, or immediately after, post a pinned comment that gives viewers an easy, low-friction reason to engage. The best pinned comments are specific questions with constrained answers:

Weak: "What did you think? Let me know in the comments!" Strong: "Which of these 5 mistakes have you made? I am guilty of number 3." Stronger: "I am curious: what is your current average view duration? Drop it below and I will reply with one tip specific to your number."

Comments in the first hours signal engagement to the algorithm. Your pinned comment can seed responses from viewers who otherwise would not have commented.

Benchmark: your pinned comment should be answerable in a few words. Long-answer prompts get fewer responses because they feel like work.

Step 25: Cards and end screen (strategic placement). Add info cards at moments when a viewer might want to go deeper on a topic you have mentioned but are not fully covering. The ideal card placement is:

• When you reference another video or resource
• When you mention a concept that has its own dedicated video
• NOT during your hook; cards here cause early exits
• NOT during your strongest content sections; do not give viewers a reason to leave during your best moments

End screen: use most of the window YouTube allows rather than a few seconds of it, and feature your best-performing related video (not your newest video; your best-performing one in a related topic).

Step 26: Captions and subtitles. Verify that YouTube's auto-generated captions are accurate. Fix any errors in YouTube Studio's caption editor. This matters because:

• Captions improve search discoverability (YouTube indexes caption text)
• Incorrect auto-captions create a poor experience and can misrepresent what you said
• Some viewers watch with captions on by default

If you have viewers in other countries (Analytics, then Audience, then Geography), consider adding translated subtitles for the languages that show up there.

Step 27: Final playback (the phone test). Before clicking publish, play the opening of your video on your phone. Not your editing monitor. Your phone. This is how most viewers will actually meet the video.

Check:

• Is text on screen readable at phone size?
• Is the thumbnail clear and compelling at phone-thumbnail size?
• Does the audio sound good through phone speakers?
• Does the hook grab you even on a small screen with potential distractions?

Benchmark: if the opening does not hold your attention on a phone screen in a well-lit room (the hardest viewing condition for contrast and detail), it will not hold your audience's attention either.

Every professional field uses checklists. Surgeons use them before operations, aviation crews use them before takeoff, and software teams use them before a release. Not because the steps are unknown, but because a routine can be executed correctly for years and still miss one step on the day it matters.

YouTube is no different. A single missing step (a poorly written title, a forgotten end screen, an unoptimized thumbnail, an opening that does not land) can sink a video that took a full day to produce. And unlike surgery or aviation, there is no co-pilot catching your mistakes. You are solo.

The problem with mental checklists is that they fail exactly when you need them most: when you are tired after a long edit session, when you are rushing to meet your upload schedule, when you are excited about a video you think is great. Excitement creates the biggest blind spots, because that is the moment you skip the checklist with "this one is different, this one is clearly good."

Two things separate this checklist from generic advice. First, every step states what passing looks like, so there is no ambiguity to rationalize your way past. Second, where a step carries a threshold, the threshold is either measured and cited or left out entirely.

The checklist is organized in production order: script, filming, editing, metadata, and upload. Each phase catches different categories of problems, and skipping any phase means accepting the risk that those problems ship.

Of the 27 steps, step 1 (script analysis) delivers the most return on time invested. Here is why:

Steps 2-27 fix execution problems: filming quality, editing tightness, metadata, publish timing. These matter, but they are optimizing the packaging. Step 1 fixes the product itself: the content structure that decides whether viewers stay or leave. Read more about this in our script analysis guide.

A concrete example. Imagine a creator who follows steps 8-27 perfectly. Their audio is pristine. Their editing is tight. Their thumbnail is optimized. Their title is compelling. But their script spends its opening on setup, has no re-engagement moment in the middle of the body, and runs on additive transitions throughout (each section feels like "here is another thing" rather than building momentum). That video will underperform despite flawless execution on everything else.

Now reverse it: a creator whose production quality is merely adequate but whose script carries the claim from the first sentence, places deliberate re-engagement moments where the body would otherwise sag, builds momentum through adversative transitions, and ends with a "one more thing" that rewards viewers who stayed. That video will outperform the first one, both in the data the audience produces and in how long the video keeps finding new viewers. Learn how to build scripts with that structure in our script structure guide.

The script is the foundation. Everything else is decoration, and you cannot decorate your way out of a structural problem. Start with step 1. Try Prepublish and run your next script through analysis before you record. The shorter version of this same check, with five tests and the evidence behind each, is check your YouTube script before recording.