Last month I opened a channel's analytics and found a video from 2023 sitting at 41,000 impressions in the previous 28 days with a 2.1% click-through rate. Nobody had touched it in two years. YouTube was still offering it to people, and the packaging was still losing the argument. That one video was quietly wasting more reach than the creator's last three uploads combined.
So the question is not whether to repackage. It is which old videos to repackage first, and in what order, because you do not have infinite designer hours and you should not spend them evenly.
Here is the ranking system I use when a channel hands me a back catalogue and asks where to start.
What makes an old video worth repackaging at all?
An old video is worth repackaging when YouTube is still showing it to people and those people are still saying no. That is the whole test, and it has two halves.
The first half is impressions. If the video is getting served, the algorithm has already decided it is relevant to somebody. You are not asking for a favour, you are asking for a better conversion on traffic you already have. The second half is click-through rate. If impressions are healthy and CTR is below your channel average, the packaging is the bottleneck, not the topic and not the edit.
Videos with almost no impressions are a different problem. New packaging on a video nobody is being offered will do very little, because there is no audience at the door to convert. That is a distribution problem, and it usually needs a fresh video on the topic instead. I wrote more about that line in when a video is actually worth repackaging.
So the short version: high impressions plus low CTR equals repackage. Low impressions plus low CTR equals leave it, or remake it.
Which videos should you repackage first?
Rank by recoverable impressions, not by how much you personally like the video. Recoverable impressions is a rough number you can calculate in about ten minutes with a spreadsheet.
Take each video's impressions over the last 90 days. Multiply by the gap between its current CTR and your channel's median CTR. That gives you the extra clicks you would win if that video simply performed like an average video on your channel. Sort descending. That sorted list is your repackaging queue.
An example with real arithmetic. Say your channel median CTR is 5.4%.
- Video A: 120,000 impressions, 2.8% CTR. Gap is 2.6 points. Recoverable clicks, about 3,120.
- Video B: 30,000 impressions, 1.9% CTR. Gap is 3.5 points. Recoverable clicks, about 1,050.
- Video C: 400,000 impressions, 5.1% CTR. Gap is 0.3 points. Recoverable clicks, about 1,200.
Most creators would fix Video B first, because 1.9% feels the most embarrassing. Video A is worth three times as much. Video C looks like the winner and has the least room, because it is already close to par.
Repackage in order of recoverable impressions, not in order of which thumbnail annoys you most. The ugliest thumbnail on your channel is often not the most expensive one.
That single reordering is usually the difference between eight thumbnails that add a few thousand views and eight that add tens of thousands. If you want the longer argument for why the back catalogue beats new production on cost per view, I laid it out in why old videos are your cheapest growth.
What about evergreen versus dated topics?
Repackage evergreen topics first, because the new packaging keeps earning for years, while a dated topic stops earning the moment the news cycle moves.
Evergreen here means the video answers a question people will still be asking in eighteen months. Tutorials, gear explainers, process breakdowns, comparisons, beginner guides. These are search and suggested workhorses. A better thumbnail on a video like that compounds, because YouTube keeps feeding it into browse and suggested surfaces without you doing anything else.
Dated topics are the opposite. A reaction to something that happened in March 2024 has a decaying impression curve no matter how good the new packaging is. You can still repackage it, but it goes at the bottom of the queue.
There is a middle category worth watching: videos that were dated when you posted them but have quietly become evergreen. A tool review that has outlived the hype cycle and now ranks for the tool's name. Those are frequently the best value on the whole list, because search traffic is patient and it keeps arriving.
How many old videos should you do in one batch?
Do eight at a time, and do not touch anything else on the channel while they run. Eight is enough to see a pattern and small enough to read the results cleanly.
That number is not arbitrary. It is roughly the point where you can tell the difference between a packaging effect and normal week to week noise. One repackaged video tells you nothing. Two tell you nothing. Eight give you a shape.
It is also why our pricing and services are built around a batch of eight custom thumbnails for $250, which works out around $31 a thumbnail, with one per video, unlimited revisions, a dedicated designer, and title consultation included. Each thumbnail takes 48 to 72 hours. A batch of eight is a normal working week, not a quarter long project.
Keep the rest of the channel steady during the batch. If you change your upload schedule, your title style and your thumbnails in the same fortnight, you will not know which change did the work.
Which videos should you skip even if the numbers look good?
Skip anything where the content itself is the reason people leave. Better packaging on a weak video buys you a worse retention curve and a slightly angrier audience.
Look at average view duration and the first thirty seconds specifically. If a video has a decent CTR and a brutal early drop, the packaging is already writing a cheque the video cannot cash. Raising CTR there makes the mismatch louder. YouTube notices, and so do viewers.
Also skip:
- Videos with outdated information that would embarrass you if a stranger found them first
- Videos whose audio quality is genuinely painful, since packaging cannot rescue that
- Videos you have already repackaged in the last 60 days, because you are still collecting data
- Sponsored videos where the sponsor relationship has ended and the pitch is now dead weight
And there is a fifth category people forget: videos where the title is the actual problem. Sometimes the image is fine and the title is doing nothing. That is why we treat packaging as the thumbnail and the title and the first frame promise as one decision, not three separate jobs. Swapping the image alone on a title problem is a wasted cycle.
What actually changes when we repackage a video?
We start from the reason the video is being offered, then design the promise around that reason.
The practical order looks like this. First, read the traffic sources. A video living on suggested needs a different promise from a video living on search. Search viewers already know what they want, so clarity beats intrigue. Suggested viewers are browsing, so the thumbnail has to create a question worth answering.
Second, decide what the single claim is. One idea, one focal point, readable at the size of a postage stamp on a phone. If I have to explain the thumbnail to you, it is too complicated.
Third, write titles against the image, not before it. A title that repeats the thumbnail text wastes half your packaging surface. The title should add the second half of the promise.
Fourth, check the first frame. When someone hovers or autoplays, the frame they land on either confirms the promise or contradicts it. A frame that contradicts your thumbnail costs you the click you just won.
We do this custom per video, never from a template, and we test styles on our own channels before they go near a client's back catalogue. If you want to see how that reasoning looks written out on real examples, the case logic behind repackaging decisions walks through the thinking.
Niche matters here too. The rules that move a fashion catalogue are not the rules that move a tutorial channel, which is why packaging for fashion creators and thumbnail design for anime channels read as different playbooks even though the underlying method is the same.
How long before you know if it worked?
Give each repackaged video 14 to 28 days and compare it against its own trailing 28 day baseline, not against your newest upload.
YouTube redistributes gradually. The impression pool for an older video does not refill overnight, and the first few days after a swap are noisy. Reading results on day three is how creators talk themselves out of changes that were working.
Compare like for like. Same video, before and after, same window length. The channel's own analytics documentation from YouTube's official help resources explains how impressions and CTR are counted, and it is worth reading once so you are not arguing with a metric you have misunderstood.
Two honest tradeoffs. First, CTR can rise while views stay flat if the video is simply not being offered much, which takes you back to the impressions test. Second, a big CTR jump on a video with weak retention can shorten its life rather than extend it. Watch both numbers together.
If you also publish long form interview content, the same batching logic applies to the edit side, which is what the sister service at Pixel Samy handles for video podcasts.
A simple order of operations
If you want the whole thing as a checklist:
- Export the last 90 days of video level impressions and CTR
- Find your median CTR
- Calculate recoverable clicks for every video
- Remove anything with a retention or accuracy problem
- Sort what is left, take the top eight
- Repackage thumbnail, title and first frame together
- Wait 14 to 28 days, compare each video to its own baseline
- Take what you learned into the next batch, and into your next upload
That last step is the one that pays twice. A repackaging batch is also the cheapest packaging research you will ever run, because you are testing ideas against traffic that already exists instead of gambling on a new upload.
If you want us to run the ranking and take the top eight off your hands, say hello and send the channel link. I will tell you honestly if I think the catalogue is worth it.