The first two weeks with a new packaging partner decide almost everything. I have watched channels sign with a studio, send over a Drive link with 400 unsorted files, and then wonder six weeks later why the thumbnails feel like someone else's channel. That is not a talent problem. That is an onboarding problem.
So here is the whole thing, written out. What we ask for, what we do with it, what you get back, and where it usually goes wrong. If you are comparing studios right now, use this as a checklist against whoever else you are talking to.
What is onboarding with a YouTube packaging agency actually for?
Onboarding exists to transfer your judgment into someone else's hands. That is the real job. You have spent years building an instinct for what your audience clicks, and most of that instinct lives in your head as unwritten rules. A good onboarding process drags those rules into the open so a designer can apply them without you in the room.
Everything else in onboarding is admin. Useful admin, but admin. The part that matters is the judgment transfer, and it is why we spend more time on your worst performing videos than your best ones. Your hits tell me what worked once. Your flops tell me where your audience stops caring, and that boundary is far more useful when I am designing thumbnail number 6 at midnight without you.
One framing note before we go further. We do not sell thumbnail design in isolation. We sell packaging, which means the thumbnail, the title, and the promise the first frame makes, designed as one unit. A brilliant thumbnail attached to a title that says something slightly different is just a fight between two pieces of your own work. If you want the longer argument for that, I laid it out in the piece on what you get with ClickTheory.
What we ask you to send before day one
We send one intake form. It is short on purpose, because a 40 question form gets abandoned and a 9 question form gets answered honestly. Here is what we want.
- Your channel link and the last 20 videos. We pull performance ourselves. You do not need to build a spreadsheet.
- Three videos you think were underrated. Good content, disappointing click rate. These are almost always the first repackaging candidates.
- Three thumbnails from anywhere on YouTube you wish were yours. Not competitors necessarily. Just images that make you stop.
- Any hard rules. No face on thumbnails. Never use red. Brand font locked. Whatever they are, tell me now rather than in revision three.
- The next 8 video topics, even if they are one line each.
- Who your viewer actually is. Two sentences beats a persona document.
- Where your face assets live. Raw stills, not previously cut out PNGs, if you have them.
- Who approves. One name. Committee approval kills packaging faster than bad design does.
That last one is worth dwelling on. When four people approve a thumbnail, you end up with the option nobody objected to, which is never the option anybody clicks. Pick one decision maker and let them decide.
How long until the first thumbnail actually lands?
The first custom thumbnail comes back in 48 to 72 hours from the point we have your intake and your assets. Not from the point you pay. That distinction matters, because the clock genuinely starts when the raw material arrives, and the fastest onboardings I have run were fast because the creator filled the form the same day.
After the first one, the rhythm settles. Each subsequent thumbnail also runs on that 48 to 72 hour window, and your dedicated designer stays the same person the whole way through, which is the entire point of a dedicated designer. By video three they stop asking you basic questions because they already know the answers.
If a video is going live sooner than that, rush is +$20 and moves you to the front of the queue. I would rather you use it occasionally on a genuinely time sensitive upload than build your whole schedule around it.
The kickoff call, and why we talk about titles first
One call, usually 30 to 40 minutes, and we spend the first half on titles rather than images. That surprises people.
Here is the reasoning. A title sets the promise. The thumbnail either reinforces that promise or complicates it, and you cannot design the reinforcement until you know the promise. If your title is a number claim, the thumbnail should not also be a number claim, because now you have said the same thing twice and wasted half your surface area. If your title is abstract, the thumbnail carries the concrete detail. Title consultation is included at no extra cost for exactly this reason. It is not an upsell, it is a prerequisite.
If your thumbnail and your title are saying the same sentence, one of them is doing nothing. Packaging works when the two halves cover different ground and land on the same promise.
On the call we also set the visual boundaries. Not a rigid template, because we do not run templates, but a set of constraints: how much of the frame your face occupies, whether text lives left or right, how many words maximum, what your background language is. Constraints make a channel recognisable in a feed. A channel with no constraints looks like eight different channels.
If you want a title pack, meaning a set of tested title options per video rather than a single consultation, that is +$25. Most creators start without it and add it around month two, once they can see which of my title instincts they trust.
What happens to your back catalogue during onboarding?
We audit it, and it is usually where the fastest wins live. Old videos are the cheapest growth most channels have, and almost nobody works them.
The logic is simple. A video published 14 months ago has already been through the expensive part. It exists, it is indexed, it has watch time history, and YouTube already knows who it suited. When it underperformed, that is very often a packaging failure rather than a content failure. Swap the thumbnail and the title, and you are asking the algorithm to reconsider a video it already has data on. That is a much shorter path than launching something new and starting from zero.
During onboarding I flag every video in your last 50 where the click rate sits well below your channel median while the average view duration sits at or above it. That combination is the tell. People who did click stayed, which means the content is fine and the packaging is what stopped everyone else. Those videos go to the top of the repackaging list.
YouTube's own guidance on thumbnails and metadata in the creator resources covers the mechanical side of swapping assets safely, and it is worth a read if you have never changed a live video's thumbnail before. The strategic side, which videos to touch and in what order, is the part we handle.
What does the revision loop look like in practice?
Revisions are unlimited, and the loop is deliberately blunt. You get a draft. You reply with what is wrong. We send the next version. Repeat until you are happy.
The thing that makes unlimited revisions work rather than turn into an endless grind is how you phrase feedback. "Make it pop" costs you a round trip. "The face is too small to read at phone size" fixes it in one. So during onboarding I ask people to give feedback in terms of what a viewer would fail to understand, not in terms of taste. Taste feedback produces a thumbnail you like. Comprehension feedback produces a thumbnail that gets clicked. Those are different objects and only one of them pays.
Average in practice is around two rounds by video four, and often one round by video six. If we are still at four rounds by video eight, that is a signal we got the brief wrong at kickoff and I will say so.
A/B variants are +$15 per video if you want a genuine second direction to test rather than a small edit of the first. I recommend them on your two or three biggest uploads a quarter, not on everything.
Should you onboard an agency or hire in house?
Hire in house when your upload volume is high enough that a designer has full time work, and when you have the management capacity to give them direction every week. Use an agency when you want a tested eye without carrying a salary, or when your volume is 4 to 12 videos a month.
I have written the honest version of that comparison in the piece on a packaging team versus in house, including the cases where hiring is genuinely the better call. The short version is that most channels under 300k are better served by an outside team for two more years than they think.
Geography does not change the answer much either, since all of this runs async. We are Dubai based and work worldwide, so whether you are a thumbnail client in London or somewhere else entirely, the process above is identical. If you are shopping around more broadly, the roundup of what makes a good packaging agency covers what to actually ask on a sales call.
What onboarding costs, and what you get
The standard pack is $250 for 8 custom thumbnails, which works out to about $31 each. One per video, unlimited revisions, a dedicated designer, and title consultation included. Every thumbnail is custom to your video. We do not run templates, and the visual approaches we bring have been tested on our own channels before they touch yours.
That pricing exists so a new channel can test us across a real run of videos rather than one lucky image. Eight videos is enough to see a pattern in your click rate. One is not. You can see the current breakdown on our pricing and services page, and a sample of finished work on our work.
One more thing worth knowing about onboarding: styles vary a lot by content type, and a talking head channel briefs very differently from a travel or lifestyle channel. If you shoot the second kind, the notes on thumbnail design for vlogs will tell you more about how we approach it than this general post can.
If you want to start, send your channel link to samy@pixelsamy.com and I will send the intake form back the same day.