People usually expect that the first thing a packaging studio does with a new channel is a big rebrand, new fonts, new colours, a new look rolled out across every video in one go, and I understand why, because that is what agencies tend to sell, but it is almost never what we do first, right, because a rebrand changes a lot of things at once and then nobody can tell which of those things actually helped, and you end up with a prettier channel and no real idea why the numbers moved or did not.
What we do instead, when a channel starts with ClickTheory, is change five things in a fairly fixed order, and the order is the whole point, since each one builds on the one before it and each one is something we can see working or not working before we move on. I am going to walk through all five the way we actually do them, so if you are thinking of hiring us, or just want to fix your own packaging, you know where the effort goes and why. The longer view of the first two weeks, the brief, the access, the first batch, is in how YouTube packaging agency onboarding goes, and this post is really the creative side of that same period.
Why not redesign the whole channel on day one?
Because if you change everything at once you lose the ability to learn anything, and the most valuable thing in the first month is finding out which single change makes the biggest difference on your specific channel.
There is also a practical reason, right, and it is about your regulars. Your audience already recognises something about your current packaging, a colour, your face in a certain position, a type style, and if all of that disappears overnight, some of your regular viewers stop recognising you in the feed, at least for a while. So we keep what is working, change what is clearly not, and move in steps that each have a reason. That is one. Secondly, a rebrand delays the first improved video by weeks, and I would much rather your next upload is packaged properly than wait for a perfect system. And on top of that, the style that suits your channel usually reveals itself after a handful of tested videos, so designing it upfront is mostly guessing.
What is the first thing you change?
The relationship between the title and the thumbnail, because on most new channels they are written separately and say the same thing twice, and fixing that pairing improves the very next video without touching the visual style at all.
This is the change I would make on almost every channel that comes to us, right, and it costs nothing in design. The creator writes a title, then someone makes a thumbnail that repeats the title in big letters over a picture, and the viewer reads the same promise twice and learns nothing extra. What works much better is when the two do different jobs, right, the thumbnail shows the moment or the result, the title adds the context or the stakes, and together they make one complete idea that neither carries alone.
So the first thing a strategist does on your channel is sit with the next video and write a set of title directions against formats that already work, before the designer touches anything. Then the designer sketches a few rough thumbnail ideas that pair with the strongest titles, and only after you pick one does anything get polished. It is the same process we run on every video afterwards, but the first time it is also a reset of how you think about the two pieces. The full argument for designing them together is in the title thumbnail pair.
Why does the face and expression come second?
Because once the title and thumbnail are doing different jobs, the face becomes the fastest thing a viewer reads, and on most channels the expression is either missing, generic, or wrong for the promise in the title.
I am not saying every thumbnail needs a face, right, some of the best ones do not, and a faceless channel has its own logic. But where the creator is the draw, the expression has to match the video. A calm explainer with a shocked face reads as clickbait, and a genuinely surprising result with a neutral face throws away the surprise. What we usually do is go back through your footage, or ask for a quick set of stills, and find expressions that fit the three or four kinds of videos you make most, so the designer has real options instead of the one frame the editor happened to export.
Size and position matter too, right, more than people think. On a lot of channels the face is too small, or pushed to the edge where the timestamp sits, or covered by text. At feed size the eyes are what people find first, so we make the face big enough that the expression reads, and we place it where nothing in the YouTube interface covers it. There is a fuller look at when faces help and when they hurt on our page about faces in YouTube thumbnails, and the logic is very very simple once you see it on your own videos.
What changes about the text, and why only third?
We cut the text down to three or four words or remove it entirely, and we do it third because text only makes sense once you know what the title says and what the face is doing, otherwise you are writing words into a frame that is still changing.
Most new channels, right, have too much text on their thumbnails, a full sentence, sometimes two, often the title again. At phone size that is unreadable, and even when it can be read, it competes with the face for attention. The question we ask is what the image cannot say on its own, and whether the title already says it. If the title covers it, the text goes. If there is one word or a number that sharpens the promise, it stays, big and with real contrast. You can see the rules we actually follow in thumbnail text best practices.
Text is also where a channel's look starts to settle. A consistent type style and a consistent place in the frame are often the first constants we keep across every video, and they do more for recognition than any colour scheme, which is why this step tends to quietly become the start of your style guide.
Why do you go back to the old videos fourth?
Because by step four we have a working approach on new uploads, and the fastest place to apply it is the handful of older videos that are still being shown to people, since improving their packaging can matter more than anything you publish next month.
Every channel with some history has a back catalogue doing quiet work, videos that still get impressions from search and suggested, still get watched, still bring in subscribers. Most of them were packaged when the channel was younger and the creator was still figuring things out, and some of them are being shown thousands of times with a thumbnail nobody would make today. Changing those, right, does not require filming anything, and the video already has watch history behind it.
We usually start with ten. We look at which older uploads still get the most impressions and which ones have a strong video underneath weak packaging, and we pick the ten where a new title and thumbnail is most likely to matter. Those get repackaged with the same process as a new upload, and where it makes sense they get an alternate version for a test, since YouTube's comparison tool works on old videos as well as new ones. The reasoning is in repackage your back catalogue before you film anything new, and it is honestly the step I would push hardest on a channel with more than a year of uploads.
Make sense so far? The first three steps are about getting the method right on one video at a time, and the fourth is where that method starts paying for the work you already did.
How does the testing habit actually work?
We ship an alternate thumbnail, and often an alternate title, alongside each video, you run it through YouTube's own test, and every result, the winner and the loser, gets written into your channel's packaging playbook so the next decision is better informed than the last.
YouTube lets creators test up to three titles and thumbnails, and at the end it shows everyone the version with the highest watch time, which is a better measure than clicks alone because it rewards packaging that matches the video. It runs from YouTube Studio on desktop, it needs advanced features, and it does not work on Shorts, made for kids videos or a few other cases, which are all set out in YouTube's guide to testing titles and thumbnails. We put testing last in the five because a test is only useful once the options being tested are genuinely good, otherwise you are just picking the less bad of two weak thumbnails.
The playbook is the part people underestimate, right, because it builds slowly. After a couple of months it holds a real record of what works on your channel specifically, which expressions, which text lengths, which title formats, which kinds of image, and that record is worth more than any general advice, including mine.
The first month is not about making your thumbnails pretty, it is about finding out what your audience actually responds to, and then doing more of that on purpose.
What do these changes cost, and what does a channel need to send?
Eight thumbnails come to 250 dollars as a pack, with title consultation and unlimited revisions included, and a test version is 15 dollars extra per video, so the first five changes cost the same as any other month of work rather than a separate onboarding fee.
What we need from you is mostly access to information. A list or a link to your recent videos, the next few you are planning, raw footage or stills for upcoming uploads, and ideally a screenshot or export of your YouTube Studio numbers so we can see which old videos still get impressions. We keep the impressions and click-through rate data in context, because a click-through rate only means something next to the number of impressions and where they came from, which YouTube explains in its guide to impressions and click-through rate. Everything runs asynchronously over Slack and email, and a normal job comes back in two to three working days, weekends excluded. The full details are on our services page.
When is ClickTheory not the right first hire?
When the videos themselves are the problem, or when the channel is so new that there is nothing yet to learn from, in which case the money is better spent on the content or on simply publishing more before paying for packaging.
I would rather say this plainly, right, than take on a channel we cannot help. If people click and then leave in the first thirty seconds, packaging will make that worse, because better thumbnails bring more people to a video that is not holding them. If you have three uploads, there is no back catalogue and very little to test against, and a solo freelancer or even a decent template can get you through the first few months more cheaply. And if you already have a designer who pairs titles and thumbnails, tests regularly and keeps a record, you are already doing what we would do, and switching just for the sake of it makes no sense.
Where we fit is a channel with some history and a steady upload rhythm, where the packaging has never been treated as its own job. The reason I trust this order is that we run it on the Pixel Samy group's own channels first, several of them past 100,000 subscribers, so the five steps are not a theory I read somewhere, they are how we package our own uploads. If you are deciding whether now is the moment, when to hire a YouTube packaging agency goes through the signals in more detail.
At the end of the day the order is the method, pairing first, then the face, then the words, then the old videos, then the habit that keeps all of it honest, and a channel that gets those five right usually finds its look on its own, without anyone having to invent one upfront. So yeah. That's my way of saying it.