The gap between 10k and 100k subscribers is where most channels quietly stall. Not because the videos get worse. Usually they get better. The problem is that the packaging that worked at 10k was carried by an audience that already knew you, and at 100k you need packaging that works on people who have never heard your name.
I run a packaging studio, so I see this from the inside. A creator books us at 14k subs with good retention, decent editing, and thumbnails they made themselves in twenty minutes because the video was already three days late. Those thumbnails were fine when 70 percent of their views came from subscribers. They fall apart the moment the browse and suggested feeds start putting that video next to a channel with a full time designer.
This post is about what actually changes across that stretch, and what I would fix first if you are sitting at 10k, 30k, or 60k and the graph has gone flat.
What actually changes between 10k and 100k subscribers?
The traffic source changes, and everything else follows from that.
At 10k subs, a large share of your views come from the subscriber feed and notifications. Those viewers click because they like you. Your thumbnail can be a mediocre frame with three words on it and the video still opens strong, because the click decision was made before the thumbnail was ever seen.
Past roughly 25k to 40k, if you are growing at all, YouTube starts testing you in browse and suggested against creators outside your subscriber base. Now the click decision is made in about one second, on a small rectangle, against eight competitors, by someone with zero loyalty to you.
That is the whole shift. You stop selling to people who already like you and start selling to strangers. Packaging is the only part of your video that talks to strangers.
Three practical consequences:
- Your CTR number stops being one number. Subscriber CTR and browse CTR diverge hard, and the average hides the problem.
- Face-only thumbnails stop carrying weight, because the stranger does not know your face yet.
- Titles have to carry the specific promise, because the thumbnail no longer gets to assume context.
Why do thumbnails that worked at 10k stop working at 50k?
Because they were never doing the work you thought they were doing.
I look at a lot of back catalogues, and the pattern is consistent. The thumbnails from the 10k era are usually inward facing. They reference a running joke, a series name, an ongoing project, or an in-joke phrase that only a regular would decode. They performed well on the subscriber feed and got read as proof the packaging was good.
Then the same style goes out to browse and the CTR drops by half, and the creator concludes the algorithm hates them or the niche is saturated. Nine times out of ten the niche is fine. The thumbnail just assumed a viewer who does not exist in that feed.
The fix is not louder. Louder is the reflex everybody has and it is usually wrong. More arrows, more red circles, more screaming face. That reads as generic, and generic loses to specific every time. The fix is to make the thumbnail legible to someone with no prior context, which usually means fewer elements, one clear subject, and a title that names the actual stake.
The channels that break 100k are almost never the ones that made the most videos. They are the ones that made their videos understandable to strangers.
Should you repackage old videos or just make new ones?
Repackage. This is the cheapest growth most channels in this band have, and almost nobody does it.
Think about what you own at 30k subs. Maybe 90 videos. Maybe fifteen of them have genuinely good retention and a real evergreen topic, and are sitting at a few thousand views because the packaging was made in a rush eighteen months ago. Those videos are finished products. The research is done, the edit is done, the hosting is free. The only broken part is the two second first impression.
A new video costs you a week and might do nothing. A repackage costs one design cycle and lands on a video YouTube already has watch time data for.
How I pick which ones to repackage:
- Sort by average view duration or average percentage viewed, not by views. You want proof the video is good.
- Filter to topics that are still true. Anything tied to a dead tool or an old year is not worth the design.
- Ignore anything with high impressions and high CTR. That one is working. Leave it alone.
- Prioritise videos with high impressions and low CTR. YouTube is offering you the traffic and the packaging is declining it.
- Change the thumbnail and the title together, and change them once, then leave it alone for two to three weeks so you can read the result.
The most common mistake is swapping the thumbnail and keeping a title that never made sense to an outsider. They are one unit. YouTube's own creator resources on how impressions and click through rate work are worth reading if you want the mechanics behind that.
If your catalogue is much larger and the channel is past six figures, the calculus shifts, and I wrote about that separately in what packaging looks like after 100k subscribers.
What does packaging mean, and why not just say thumbnails?
Packaging means the thumbnail, the title, and the first frame promise, designed as one decision.
I am a little tired of the word thumbnail, honestly. It describes a file, not a job. The job is: in one second, on a phone, make a stranger understand what they get and why it matters. A thumbnail is one input to that. The title carries specificity that an image cannot. The first fifteen seconds of the video have to pay off what the pair of them promised, or your retention collapses and the algorithm stops distributing you regardless of how good the click rate was.
This is why we do title consultation as part of every package rather than selling it as a design service. Handing back an image and letting you attach whatever title you already wrote is how you get a beautiful thumbnail that underperforms.
It is also why we do not use templates. A template is a decision made before knowing the video. The whole point of packaging in this subscriber band is that the video is competing against specific other videos, in a specific feed, for a specific viewer state. You can see how this plays out across different niches in our work.
How many thumbnails do you actually need to test?
More than one per video, fewer than you think per month.
Here is the honest tradeoff. Testing two variants on a video tells you something real, but it costs you a design cycle and it splits your data. On a channel doing 5,000 impressions per video, a variant test will take a long time to say anything statistically meaningful. On a channel doing 200,000 impressions, it resolves in a day.
So my rule for this band is roughly:
- Under 30k subs: one strong thumbnail per video, and put your testing energy into the repackaging queue instead. You learn faster from ten repackages than from two A/B tests.
- 30k to 100k: variant test your biggest swings only. The video you think could break out. Not the routine uploads.
- Past 100k: test more or less everything, because impressions arrive fast enough to make it cheap.
That is why our A/B variant sits as a $15 add on rather than being baked into the base price. Most channels in this range should not be paying for it on every video.
If you are changing direction as well as scaling, the sequencing gets trickier, and repackaging through a niche change has its own rules that override some of this.
What should you spend on packaging at this stage?
Enough that it stops being the bottleneck, and not a dollar more.
Our base is $250 for 8 custom thumbnails, which works out around $31 each. One per video, unlimited revisions, a dedicated designer who actually learns your channel rather than a rotating queue, and title consultation included. Turnaround is 48 to 72 hours per thumbnail. Add ons if you want them: an A/B variant at $15, a title pack at $25, and rush at $20.
I put the numbers here because I think this decision should be boring. Eight thumbnails is roughly two months of uploads for a weekly channel, or one month plus a repackaging batch. If a single repackaged video recovers a few thousand extra views on an evergreen topic, the batch has paid for itself, and that is before the new uploads.
What I would not do is hire a full time designer at this stage. The volume does not justify it, and you will spend more time art directing than making videos. Full pricing and scope is on pricing and services.
Does the niche change how packaging should look?
Yes, and more than most advice admits.
Face driven niches and object driven niches behave differently. A sports channel is competing in a feed full of high motion imagery and known logos, which changes both the crop and the contrast decisions, and I went into that in detail in packaging for sports channels. A coaching or education channel is selling an outcome that does not photograph, so the thumbnail has to visualise a before and after state that does not literally exist, which is a different craft entirely.
Founder led channels are their own case. If you are building a personal brand around a product, the thumbnail has a second job beyond the click, which is establishing that you are a credible person to learn from. That constraint shapes everything, and it is why thumbnail design for SaaS founders is a different conversation from generic creator packaging.
We test styles on our own channels first before we put them on a client channel. That is not a marketing line, it is just cheaper to be wrong on our own inventory.
What does the first month with a packaging studio look like?
Mostly it looks like sorting your back catalogue, which is the least glamorous and most valuable part.
Week one we go through your analytics with you and build the repackaging queue: the videos with proven retention and bad packaging. We usually find between five and twelve on a channel of this size. Week two the first thumbnails land, split between new uploads and the queue. From there it is a rhythm, and the interesting work becomes reading the results and adjusting the visual language rather than starting from zero every time.
We are Dubai based and work worldwide and async, which in practice means you brief in your evening and see work in your morning. Fine detail: we never show client names publicly, so if you are wondering why there are no logos on this site, that is why. We will talk about a finance channel around 80k subs, never who it was.
Some creators come to us through the video side of the business, since my team also edits video podcasts, and packaging is usually the piece that was missing there too.
The short version
If you are between 10k and 100k, your videos are probably better than your packaging, and the fastest available growth is sitting in videos you already published. Fix the packaging for strangers, treat the title and the image as one object, and work the back catalogue before you work harder on new uploads.
If you want a second pair of eyes on your catalogue before deciding anything, contact me and I will tell you honestly whether packaging is your bottleneck or not.