We turned down four projects last month. One of them was worth more than everything else we booked that week. I still think it was the right call, and I want to explain why, because "what we say no to" is the only part of running a small studio that nobody puts on a website.
Most studios advertise what they do. Fair enough, that is what buys groceries. But the shape of a studio is decided by the work it refuses. Every yes eats a slot. We ship one pack at a time, $250 for 8 custom thumbnails, one per video, unlimited revisions, a dedicated designer on it. That is roughly $31 a thumbnail and 48 to 72 hours per thumbnail of actual attention. There is no version of that math where we can take everything and still be good.
So here is the honest list. Not a manifesto. Just the things that come into the inbox and get a polite no, and what each one taught me.
No to "just make it look nice"
The most common request I decline is the one that sounds easiest. Someone sends eight video files and says make the thumbnails look good, no other context, no titles, no sense of who the video is for.
I cannot do that job well. A thumbnail is not decoration, it is a claim. It says here is the specific thing you get if you spend twelve minutes with me. Without a title, without a channel, without knowing whether the viewer is browsing home or searching a problem, "looks nice" means "looks like everything else in the feed."
This is why title consultation is baked into the pack instead of sold separately. The title and the image are one unit. If the title carries the number, the image should carry the face and the stakes. If the title is vague and moody, the image has to do all the informational work. I have watched a good image lose to a mediocre one because the mediocre one was arguing with its title instead of repeating it.
A thumbnail that agrees with the title twice is a wasted slot. Say two things, not one thing louder.
No to template packs
We get asked for template systems maybe twice a month. Buy one design, get a Photoshop file with swappable text, produce fifty thumbnails yourself. Cheaper per unit, faster, feels scalable.
I will not sell that, and it is not precious of me. It is that templates fail in exactly the place they promise to help. The value of a thumbnail comes from difference, and a template is a machine for producing sameness. Week three of a template looks fine. Week nine looks like a channel that stopped trying. Viewers do not consciously register it. They just scroll a little faster past your row.
Every thumbnail we make is custom per video. That is slower and it is the whole product. If you want the honest comparison of doing this in house versus handing it out, I wrote about where in-house design actually beats an outside studio and it is not a clean win for us in every case.
No to unlimited retainers with an unlimited queue
Design subscriptions look great on the pricing page. One flat monthly fee, unlimited requests, one active at a time. I understand the appeal completely and I have thought hard about launching one.
The problem is the queue. Unlimited requests plus one active slot means the constraint moves from your budget to your patience, and the studio's incentive quietly flips. Fast shallow work clears the queue. Careful work clogs it. I have seen what that does to quality over six months, and I broke down the mechanics in why design subscriptions rarely stay a good deal.
A pack of 8 has a floor and a ceiling. You know what you paid, I know what I owe, and unlimited revisions applies to the eight we agreed on rather than to an infinite backlog. Revisions should be unlimited. Scope should not.
No to work I cannot test
We test styles on our own channels before we put them on yours. That sounds like a nice line until you realise it rules things out.
Somebody asks for a style that is trending on a channel with two million subscribers. Extreme compression, three words, an expression cranked to eleven. It works there because the channel has authority and the audience already knows the host's face. Ported onto a channel at 20k subs, the same treatment reads as noise. The face means nothing yet. The three words have no context to lean on.
I would rather say no than run an untested import on someone's real upload schedule. The one exception is a paid variant, the A/B add-on at $15, where we ship two and let the impressions decide. That is a test with a control. Copying a big channel and hoping is not.
The most useful thing we ever learned this way came from a face experiment across a batch of our own uploads. Removing the face from tutorial style content did not tank clicks the way I expected, and adding one to a listicle helped more than I expected. I put the actual numbers and the caveats in what changed when we tested faces, including the part where the result did not hold on one category at all.
No to clients who want to name-drop
Every few weeks someone asks for a logo wall, or asks whether we can say publicly that we work with them, or asks who else we have worked with by name. The answer to all three is no.
We do not publish client names. Not on the site, not in a case study, not in a pitch. A creator's thumbnail strategy is competitive information, and a channel that pays for outside design usually does not want that on a slide. So when I talk about results, I talk about niche and subscriber band. A finance channel around 80k subs. A crafts channel in the low tens of thousands. That is as specific as it gets.
It costs us sales. Social proof sells, and I am choosing to sell with the work instead, which is why our work is the page I actually care about. If the images are not convincing, a list of names would not have saved it.
No to the wrong niche at the wrong moment
I turn down projects where I do not think we will be the ones who move the number. Some categories are genuinely hard for us. Some are hard for everyone, and the honest answer is that the format is the problem rather than the art.
Where we do best is content where the payoff is visual and the promise is concrete. Craft and DIY is the clearest case. The object is the hook, the before state is the tension, and the copy can get out of the way. We have a whole approach for thumbnails for DIY and crafts channels that leans on the finished object more than on the creator's face.
The mirror image of that is text-heavy design, which is where most amateur thumbnails die. Too many words at too small a size. Our rules for that live in how thumbnail text actually behaves at small sizes, and honestly, the single fastest improvement most channels can make is cutting the copy in half and doubling the size of what remains.
Related failure, and I am guilty of it too: designing at 100 percent zoom on a big monitor. It flatters everything. Most of the audience sees the image at the width of a business card. I caught myself doing it on a batch I had already approved, which is the whole reason we check every file at feed size before it ships.
No to rush as the default
Rush exists. It is $20 and it is real, and sometimes a video has to go out Friday. What I say no to is rush as the standing operating mode.
48 to 72 hours per thumbnail is not us being slow. It is the time it takes to try a version, dislike it, and try the version that actually works. The first idea is usually the obvious one, and the obvious one is the one three other channels in your niche already shipped this week. The gap between draft two and draft four is where the click lives.
If your pipeline structurally cannot give a designer two days, the fix is upstream in your production schedule, not in the design brief. Booking thumbnails a week behind your edit is the cheapest change available and it costs nothing.
What the noes add up to
I have said no to reliability problems too, which is a strange thing to admit as someone selling a service. Cheap and fast is a real option and plenty of creators run on freelancers successfully. But the failure mode is specific and it is not about talent, it is about the week they go quiet. I laid out the freelancer reliability problem without pretending a studio is automatically the answer.
We are a small operation in Dubai working async with people all over the world. The sister side of the business edits video podcasts and clips from $199, which is a useful pressure valve, because when someone needs something we are not good at I can sometimes point them at a room where the answer exists.
The whole point of a short list of noes is that it makes the yes mean something. When I take a pack, it is because I think eight custom thumbnails, tested, revised until you stop flinching at them, will change the shape of the channel's next quarter. If I do not think that, the honest move is to say so on the first email rather than on the fourth revision.
If you are weighing whether your channel is one we would take, the fastest way to find out is to ask. Send the last eight titles and I will tell you straight, including if the answer is no.