Last month I opened a channel's analytics and found a video from 2023 that had done 41,000 views in its first two weeks and then flatlined for eighteen months. The creator wanted to delete it. Same week, another creator asked me whether he should "clean up" 60 old uploads because they were dragging his average. Both of them were about to throw away the cheapest growth they had.
So let me answer the question properly, because repackaging old videos vs deleting them is a real fork in the road and the wrong turn costs you a year.
Does deleting old videos actually help your channel?
Almost never, and not in the way people think. The idea that a weak old upload "poisons" your channel comes from a misreading of how the system works. YouTube does not average your channel into one score and then hand you reach based on it. Recommendations run per video, per viewer, per session. A quiet video from 2023 is not standing in front of your new upload with its arms out.
What deletion actually does is permanent. You lose the watch time, the comment history, the accumulated search authority, the embeds, and any external traffic the video still earns. You also lose the one asset that matters most for repackaging: a video that already proved a topic works.
If you want the platform's own framing on how content gets surfaced and what genuinely affects distribution, the official YouTube creator resources are worth reading before you delete anything permanent.
There are three cases where deleting is fine. Content that breaks a guideline. Content that names a client, a price, or a claim you can no longer stand behind. Content that is factually wrong in a way that damages you. That is the whole list. "It looks bad next to my new stuff" is not on it.
What is repackaging, and is it just a new thumbnail?
Repackaging is redesigning the thumbnail, the title, and the first frame promise as one unit so the video gets a genuine second launch. It is not just a new thumbnail. That is the old framing, and it is why so many people try it once, swap an image, see nothing move, and conclude repackaging does not work.
Here is the mechanic. A thumbnail buys the eye. The title sets the specific promise. The first ten seconds either confirm that promise or break it. If you change only the image, you have made a louder version of the same offer. Click through rate might twitch. Retention will not, and the platform stops feeding the video within a day or two because the session outcome did not improve.
Real packaging work asks a harder question: what is the single most interesting thing this video actually delivers, and is that thing currently visible in the frame and in the words? Nine times out of ten it is buried in minute seven, and the thumbnail is advertising something generic instead.
The video is rarely the problem. The promise on the outside of the video usually is.
We treat all three surfaces as one design decision. That is the whole thesis behind everything we build, and the reason we call it packaging rather than thumbnail design.
Which old videos are worth repackaging?
Videos where the topic clearly landed but the packaging did not. That is the filter, and it is more specific than "anything with low views".
Here is how I sort a back catalogue:
- Strong retention, weak click through rate. This is the gold. People who found it stayed. Almost nobody chose it. Repackaging these is close to free views.
- Strong impressions, weak click through rate. YouTube offered it repeatedly and viewers said no. The offer is wrong, not the topic.
- Evergreen topic, dated framing. The advice still holds but the title reads like 2022. New title, new frame, new life.
- Weak retention, weak everything. Leave it. Repackaging a video people actively leave is putting a new door on an empty room.
- Sitting on a search term. If it still pulls impressions from search, that video is an asset. Never delete it.
I go deeper into the sorting logic in the piece on when to repackage an old video, including the retention thresholds I use as a rough cut. If you would rather someone else run that pass across your whole library, that is exactly what a back catalogue audit is for.
How much lift should you honestly expect?
Enough to matter, not enough to justify hype. I am not going to give you a magic multiplier, because anyone quoting a fixed percentage across all channels is guessing.
What I can tell you is the mechanic. Click through rate is a multiplier on impressions you are already earning. If a video is showing to people and converting at a low rate, moving that rate has compounding effects, because better performance in one cycle earns more impressions in the next. That loop is why packaging work on an existing video is so much more efficient than shooting a new one. The distribution already exists. You are changing the conversion on it.
The honest tradeoff: the loop runs both ways. A repackage that performs worse gets fewer impressions, and you can sit below where you started for a couple of weeks. That is why I always ask creators to keep the original files, and why I prefer testing one variant at a time rather than redesigning thirty videos in a weekend.
For a sense of what normal actually looks like by niche and format, so you are not measuring yourself against a number that was never realistic, have a look at the old video CTR benchmarks I put together.
Repackaging vs making a new video: which is the better spend?
Repackaging usually wins on cost per view, and it is not close for most channels.
Run the numbers on your own situation. A new video costs you scripting, a shoot day, editing, and thumbnail work, and then it has to earn its audience from zero. A repackage costs one design cycle and inherits every impression the video has already accumulated. For a channel with 40 or 50 uploads sitting in the archive, that archive is a bigger opportunity than the next three uploads combined.
Our pricing makes this concrete. A pack of 8 custom thumbnails is $250, which works out to about $31 each, one per video, with unlimited revisions and a dedicated designer. Title consultation is included, because a thumbnail without a matching title is half a job. Turnaround is 48 to 72 hours per thumbnail. If you want a second variant to test against, that is $15. A title pack is $25. Rush is $20.
At about $31 a video, eight repackages cost less than most people spend on a single new upload. That is the entire argument.
The full breakdown of what sits inside a pack is on the pricing and services page.
What actually changes in a repackage?
Concretely, four things, and they get decided together rather than in sequence.
The subject. Most old thumbnails show a wide shot of a person or a generic object. We pick the one visual that is specific to this video and nothing else. Specificity stops a scroll. Brightness does not.
The face and the framing. If you are on the thumbnail, the crop, the eyeline, and the lighting are doing more work than the text. Most creators shoot these as an afterthought at the end of a tired filming day. I put the full method in the guide on how to photograph yourself for thumbnails, and it is the highest leverage thing you can fix on your own this week.
The words. Three to five words on the image maximum, and they must not repeat the title. The image and the title should be two halves of one sentence, not the same sentence twice.
The first frame. The moment the click lands, the video has to confirm the promise. If your repackaged title promises a specific number and the video opens with a 20 second intro animation, you have engineered a bounce.
Every style we use gets tested on our own channels before it goes near a client library. We do not run templates. Each one is built for the specific video, which is also why I refuse to quote a fixed lift figure.
How do you choose who does this work?
Ask for the reasoning, not the portfolio. Anyone can show you nice images. Very few can tell you why a particular video needed that particular promise, and what they would test next if it underperformed.
The questions I would put to any studio, including mine:
- Do you change the title as well as the image, or only the image?
- What happens if the repackage performs worse than the original?
- How many revisions before it costs extra?
- Do I work with one designer or a rotating pool?
- Can you talk me through a video that did not work?
I expanded that list in the post on questions to ask a YouTube packaging agency. Use it on me too. I would rather answer the awkward version now than a month in.
We run from Dubai, work async with creators worldwide, and never show client names anywhere, which is why you will see phrases like "a finance channel around 80k subs" instead of logos. If you also make long form interview content, the sister studio at Pixel Samy edits video podcasts.
So, repackage or delete?
Repackage, unless the video is broken in a way design cannot fix.
Delete for policy, accuracy, or credibility reasons. Keep and repackage for everything else. Start with the five or six videos that have decent retention and weak click through rate, give each one a real second launch with a new thumbnail, a new title, and a first frame that matches, then watch for two weeks before touching anything else.
If you want a second opinion on which videos in your library are worth the effort, send me the channel link at samy@pixelsamy.com or use the contact form and I will tell you honestly whether there is anything there.