I have watched creators spend six weeks on a video and eleven minutes on the packaging. Six weeks of scripting, shooting, re-shooting the intro, colour grading, sourcing music. Then, at 1am the night before, a thumbnail gets exported from a phone app and a title gets typed into the upload box because the upload box was there.
That last eleven minutes decides whether the six weeks gets seen.
So this is the packaging checklist we actually run before we let anything go live. Not a vibes list. A sequence of checks with a reason attached to each one, in the order we run them, with the arguments we have had with ourselves about each. Steal all of it.
What does a packaging checklist actually cover?
A packaging checklist covers three things that ship together: the thumbnail, the title, and the first frame of the video. Not the thumbnail alone. That is the old framing and it is why so many channels feel stuck.
Packaging is the promise. The thumbnail makes the promise visual, the title makes it specific, and the first eight seconds either pay it off or burn the click. If those three are made by three different people at three different times, they will disagree, and the viewer feels the disagreement as "this is not what I thought it was" and leaves at forty seconds.
So every check below is scored against the same question: does this make one clear promise that the video keeps?
The pre publish checklist, in running order
1. Say the promise out loud in one sentence. Before you look at a single pixel. If you cannot say it in one breath, the packaging cannot carry it. "I bought the cheapest and the most expensive version of the same thing and one of them broke." That is a promise. "My thoughts on gear" is not.
2. Write eight titles, not one. Eight is the number where you stop writing the obvious one. Titles one through three are what you already thought. Four and five are worse. Six through eight are where the interesting angle shows up because you have exhausted the safe options.
3. Pick the noun that carries the curiosity. Most titles have one word doing the work. Find it. If nothing is doing the work, the title is a description, not a title.
4. Check the title at truncation length. Roughly forty five to fifty characters survive on mobile home feed before the cut. Front load. If your hook is at the end, the hook does not exist for most of your audience.
5. Thumbnail at 20 percent size. Shrink it to the size of a mobile browse card and look at it from an arm's length away. If you cannot tell what it is in half a second, nothing else on this list matters.
6. Count the readable elements. Three or fewer. A subject, one text block, one context element. Four is usually the point where a thumbnail turns into a poster and a poster does not survive being shrunk.
7. Text under four words. Ideally three. The thumbnail text should not repeat the title, it should complete it. Title says the what, thumbnail says the twist, or the other way round. Repetition wastes half your surface area.
8. Face check. Is the expression legible at card size, and is it the right expression for the promise? When we ran our own face versus no face experiments the finding was not "faces win", it was that a specific, readable emotion wins and a generic smiling head loses to a clean object shot.
9. Contrast against both feeds. Your thumbnail sits on white for some viewers and near black for others, and the near black group keeps growing. Dark edges vanish on dark mode. We wrote up the full set of adjustments for designing thumbnails that survive dark mode because it is the check most people skip entirely.
10. Sit it next to the competition. Search your own keyword, screenshot the results page, and paste your thumbnail into that grid. You are not designing a thumbnail, you are designing the odd one out in a grid of twenty. If yours matches the row, redesign it.
11. Check it against your own last six. Different enough to feel new, consistent enough to feel like you. Channels lose more to sameness than to ugliness.
12. First frame agrees with the thumbnail. Open the video and look at second zero. If the thumbnail promised a wrecked kitchen and the first frame is you sitting in a chair saying "hey guys", you just taxed your own retention.
13. Sleep on it, or hand it to someone who did not make it. Ten minutes of a fresh pair of eyes beats two hours of your own.
The best packaging decision most channels can make this month is not a better thumbnail on the next video. It is a better thumbnail on a video that already has watch time and is still being recommended.
Why is the title part of a thumbnail checklist?
Because YouTube shows them together and viewers read them together, so testing them apart tells you almost nothing. A thumbnail that fails with one title can win with another, and the reverse is just as common.
We do title consultation with every packaging job for exactly this reason. If a designer makes a thumbnail without knowing the title, they will fill the empty space with words, and those words will fight the title once it exists. Two competing hooks is worse than one clear one.
The practical version: write the title before the design starts, then let the design change the title. That loop is the work. When we deliver a set of eight, the titles usually shift two or three times as the visuals reveal a stronger angle.
When should the checklist run?
The checklist should run days before upload, not on upload day. That is the single change that improves output quality more than any design skill.
Packaging made at midnight is packaging made without options. There is no time to look at it fresh, no time to try the second concept, no time to run the truncation check properly. You ship the first idea because it is the only idea.
We push clients toward batching for this reason, and we wrote out the whole system in our piece on scheduling thumbnails ahead of your upload dates. Our turnaround is 48 to 72 hours per thumbnail, so an eight video batch slots comfortably into a month of uploads if you plan the calendar first. Rush exists as a $20 add on and we would rather you never needed it.
What should you check on old videos?
On old videos, check whether the packaging matches what the video actually turned out to be, because a lot of back catalogue was packaged before anyone knew what the video was worth.
This is the cheapest growth most channels have and almost nobody does it. You have videos that already carry watch time, already have session data, already sit in suggested feeds. New packaging on those does not need to win a cold start. It only needs to convert a little better on impressions the algorithm is already handing you.
How we pick which old videos to repackage:
- High impressions and low click through. The system is offering it and viewers are refusing.
- Good average view duration and low views. The video is strong, the packaging is not selling it.
- Anything evergreen that still gets search traffic. Those compound.
- Anything where your channel style has moved on so far that the old thumbnail looks like a different channel.
Leave alone anything that is currently performing above your channel average. Do not fix a working promise.
Google's own YouTube help documentation covers the mechanics of swapping a thumbnail on a published video, and it is worth reading the section on custom thumbnail requirements before you queue a batch of changes.
How many revisions should this checklist need?
If the brief was good, most thumbnails land in one or two rounds, and the ones that need five are usually a briefing problem rather than a design problem. Revisions are cheap. Restarts are expensive. The checklist above exists to catch the restarts before anyone opens a design file.
We do not cap revisions, and we say that because the alternative encourages designers to defend weak work instead of fixing it. But unlimited does not mean infinite in practice. We wrote about the normal revision counts we actually see if you want the honest numbers rather than the marketing version.
The thing that reduces rounds most: send the video, or at least the script, not just the title. A designer working from one line of text is guessing at the promise. A designer who has seen the footage is choosing between real moments.
What does a good brief look like?
A good brief is four lines long and includes the promise, the audience, the one visual you already have in your head, and the thing you do not want.
Here is the shape we ask for:
- Promise: the one sentence from step one.
- Who it is for: beginner, intermediate, or someone who already knows the topic. This changes vocabulary and how much explaining the thumbnail has to do.
- Assets: any footage, product shot, or screenshot that has to appear.
- The veto: what would make you reject it. "No red arrows" or "I do not want my face on this one" saves everyone a round.
That is it. Longer briefs are not better briefs. They are usually one person trying to art direct through text, and that is what the revision loop is for.
Where a studio changes the maths
Custom work per video, never templates, is the part that a checklist alone cannot give you. A template survives one upload and then becomes wallpaper on your own channel.
Our packaging batch is $250 for eight custom thumbnails, about $31 each, one per video, with a dedicated designer, title consultation, and unlimited revisions. Add ons are simple: an A/B variant for $15, a title pack for $25, rush for $20. We test styles on our own channels before we put them on yours, which is why our recommendations come with reasoning rather than trends.
We work remotely from Dubai with creators worldwide and async, which suits this workflow well, and you can see how that plays out for creators in specific markets in our notes on working as a YouTube thumbnail designer for London channels. The full breakdown of what is included sits on our pricing and services page.
If you also publish long interview content, packaging and edit are the same conversation, and our sister service handles video podcast editing on the same principles.
Run the checklist on your next upload, then run it backwards across your last twenty videos. If you would rather hand the whole thing over, email samy@pixelsamy.com and tell me what you are publishing next month.