A creator sent me a screenshot last month with one line under it: "my best video this week is from 2023." The video was 26 months old. It was pulling more daily views than his last four uploads combined, and he was genuinely annoyed about it.
I understand the annoyance. You get better at this. Your audio is cleaner, your edit is tighter, your writing is sharper, and then some scruffy old upload with a thumbnail you made in twenty minutes keeps eating your lunch. But that video is not mocking you. It is telling you something specific about what your audience wants and how they decide to click, and most channels throw that information straight in the bin.
Why do old videos outperform new ones on YouTube?
Almost always because the old video won the packaging and topic lottery, and the new ones did not. Everything downstream, the browse impressions, the suggested placements, the search ranking, follows from that.
Here is the mechanic in plain terms. YouTube shows a video to a small pool of people. If enough of them click and then stay, it shows it to a bigger pool. That loop does not care how old the file is. A video from 2023 that earns clicks and watch time today gets fed today, and a video from last Tuesday that does not, does not. Age is not a penalty. Performance is the only currency.
So when an old video is winning, one of four things is usually true:
- The topic has evergreen demand. People search or get suggested it constantly, and the video answers a question that never expired.
- The packaging is stronger than you remember. The old thumbnail was blunt, high contrast, and readable at phone size. Your new ones got prettier and less clickable.
- It has accumulated signals. Two years of watch time, comments, and end screen traffic give it a head start every time it competes for an impression.
- Your new videos are drifting. You are making what you find interesting now, not what the audience showed up for.
Most channels have a mix of all four. The good news is that three of them are fixable in a weekend.
How do I tell if it is the packaging or the topic?
Compare click through rate and average view duration side by side, and let the gap tell you which lever to pull.
Open YouTube Studio, go to the advanced mode of your analytics, set the range to the last 28 days, and sort by views. Pull the old winner and your three most recent uploads into the same table. Then read it like this:
High CTR on the old video, low CTR on the new ones, similar retention. This is a packaging problem, full stop. The audience wants both topics but only one of them looks worth clicking. Fix the new thumbnails and titles.
Similar CTR, much better retention on the old video. This is a topic and delivery problem. The click is fine, the payoff is not. Packaging will not save you here, the content has to change.
High CTR and high retention on the old video, and it is still climbing. This is the best case and the one people ignore. You have a proven vein. Make two or three more videos adjacent to it, and repackage the older cousins of that video that never got their shot.
Everything is flat except impressions. You are probably being served to the wrong audience, which is usually a title and thumbnail mismatch problem again. Google's own documentation on how the systems surface content is worth reading once properly at the YouTube Help resources, because a lot of creator folklore about "the algorithm punishing old uploads" simply is not true.
Write the answer down before you touch anything. Half the mistakes I see come from creators redesigning thumbnails when the real issue was that the video did not deliver what the title promised.
Your old winner is not a fluke to be resented. It is the only piece of free market research your channel has ever produced, already paid for, already validated by real clicks.
Should I repackage the old video that is already winning?
Usually no, or at least not first. Do not touch a video that is actively climbing. Touch the twenty videos sitting next to it that never got seen.
This is the single most common mistake. A video is finally working, so the creator immediately swaps its thumbnail to "push it further" and kills a curve that took months to build. The uplift you might gain is small, the downside is real, and you have just thrown away a control.
The money is in the neighbours. If a video about one specific topic is pulling, you almost certainly have three or four videos on closely related topics that were published with weak packaging, buried in week one, and never resurfaced. Those are the ones to fix. They already have the content quality. They just never got a fair audition.
That is what a proper back catalogue refresh of titles and thumbnails does. You are not editing the videos. You are giving thirty months of work one more chance to be clicked, at a fraction of the cost of producing anything new.
What does a repackaging pass actually involve?
A repackaging pass means rewriting the title, redesigning the thumbnail, and aligning the first eight seconds so the promise holds. All three, together, or you have not repackaged anything.
This is why I keep saying the job is packaging, not thumbnails. A thumbnail on its own is a picture. Packaging is a promise made by the thumbnail and the title at the same time, then honoured by the opening of the video. When those three disagree, retention collapses and YouTube stops serving you no matter how nice the artwork is.
Here is the sequence I use on a back catalogue:
- Rank the catalogue by rescue potential. Good retention, bad CTR, evergreen topic. Those go first. High CTR and terrible retention are a different, slower fix.
- Rewrite the title around the actual question the viewer has. Old titles are usually clever. Clever loses to clear.
- Design one thumbnail per video, from scratch. Not a template with the text swapped. The whole point is that each video has a different hook.
- Check the first frame and the first line of the video. If the thumbnail says one thing and the intro rambles for thirty seconds before getting there, the click was wasted.
- Change one video at a time and give it fourteen days. Batch swapping everything on a Monday means you learn nothing.
We walk through the full sequence in more detail in what a packaging agency actually does, including how we decide which videos to leave alone.
How many old videos are worth fixing?
Fewer than you think. On most channels the honest answer is somewhere between six and fifteen videos, not the whole library.
There is a temptation to redo everything. Resist it. A five year old video on a topic nobody searches anymore does not become valuable because it has a nicer thumbnail. You are looking for a specific profile: the video is still getting impressions, the retention is decent, and the topic still matters. Everything else is nostalgia.
This is exactly why our core package is eight thumbnails for $250, which works out around $31 each, one per video. Eight is usually the right size for a first repackaging sprint. It is enough to cover a real cluster of related videos, small enough that you can read the results, and cheap enough that one recovered video pays for the whole batch. Every thumbnail is custom for that specific video, with unlimited revisions and a dedicated designer, and title consultation is included because the title is half the package. Turnaround is 48 to 72 hours per thumbnail.
If you want to test two directions on the same video, an A/B variant is $15. A written title pack is $25. Rush is $20. That is the whole menu, and the full breakdown lives on our pricing and services page.
For channels with a genuinely large dormant library, the sprint approach and the sequencing matter more than the volume. I have written separately about reviving a dormant channel because the psychology of restarting is a different problem from optimising an active one.
What should I change about the new videos?
Stop treating packaging as the last step. Decide the title and the thumbnail concept before you record, and let the video serve the promise instead of the other way around.
The reason your new uploads underperform is rarely that they are worse videos. It is that the old winner had a sharp, accidental promise and the new ones have careful, thoughtful, forgettable ones. You got more sophisticated and less legible.
Three practical corrections:
- Write the title first. If you cannot write a title you would click, do not shoot the video yet.
- Design at phone size. Most of your impressions are 380 pixels wide. If the thumbnail needs squinting, it fails, no matter how good it looks on your monitor.
- Limit the idea to one. One face, one object, one line of text, one clear tension. Two ideas in a thumbnail equals zero.
If you want to go deeper on the craft side, the YouTube creator resources cover the fundamentals well, and the analytics guidance there is more current than most of the third party advice floating around.
When is it worth handing this to someone else?
When the videos are good and the clicks are not, and when you have stopped being able to see your own thumbnails objectively.
That second one is real. After eighty uploads you cannot look at your own work cold anymore. You know what the video is about, so the thumbnail always makes sense to you. The stranger scrolling at 11pm has none of that context.
The other trigger is opportunity cost. If repackaging fifteen videos takes you three weekends you were going to spend writing, hiring is just cheaper arithmetic. We are Dubai based and work worldwide and async, so the back and forth happens in your inbox rather than on calls. Styles get tested on our own channels before they go anywhere near a client, and we never show client names, so if you see us reference a finance channel around 80k subs, that is deliberately as specific as we get.
I have written a longer honest checklist on when to hire a packaging agency, including the cases where you should absolutely keep doing it yourself. If you already know you want the back catalogue handled, our YouTube video packaging service page explains how the eight video sprint runs, and our work shows the range of styles.
The uncomfortable conclusion
Your old videos outperforming your new ones is not a sign that you peaked. It is a sign that you have unrealised inventory sitting in your library and a packaging habit that got worse as your production got better.
Most channels can double their monthly views without filming anything new. Not because repackaging is magic, but because a video nobody clicked is a video nobody watched, and that has almost nothing to do with how good the video was. Fix the promise, and the work you already did starts earning again.
If you want a second pair of eyes on which eight videos to start with, send me the channel at samy@pixelsamy.com and I will tell you honestly whether it is worth it.