I once counted the revision notes on a single thumbnail job. Nine comments. Seven of them were opinions about color. Zero of them said what the video was actually promising the viewer. That thumbnail went through four rounds and ended up almost identical to version one, just bluer.
That is the normal failure mode. Not bad designers, not lazy creators. Just feedback that describes taste instead of describing the job.
Good thumbnail feedback is a skill, and it is learnable in about ten minutes. Here is how I ask the creators I work with to do it, and why it usually turns three rounds into one.
Why does most thumbnail feedback go nowhere?
Because it comments on the picture instead of the promise. A thumbnail is not a piece of art you are approving. It is a claim you are making to a stranger who is scrolling, and the title is making that claim with it. When your note is "I don't love the green", your designer has nothing to act on except swapping a hex code and hoping.
There are three things a viewer decides in the half second your thumbnail gets:
- What is this about
- Why should I care right now
- Is this worth my next eight minutes
Almost every useful note maps to one of those three. Almost every useless note maps to none of them. If you catch yourself writing a comment that does not touch readability, clarity of promise, or curiosity, it is probably a preference, and preferences belong at the very end of the process, not the start.
The other reason feedback stalls is that people review thumbnails at the wrong size. You are looking at a 1280 by 720 file on a 27 inch monitor. Your audience is seeing something the width of a postage stamp on a phone, or a slightly bigger rectangle on a TV with the title stacked under it. Half of the notes I get would disappear if the creator had shrunk the image to 20 percent first.
The three layer method for thumbnail notes
I ask every creator to give feedback in three passes, in this order. Do not mix them. The order matters because fixing layer one often deletes the need for layers two and three.
Layer 1, the promise. Does the image plus the title say the same thing? Is that thing the actual most interesting part of the video? This is the layer where a note is worth hundreds of views. Example: "The thumbnail says this is about budgeting. The video is really about why my old budget failed. Push the failure."
Layer 2, legibility. At thumbnail size, in a crowded feed, what reads and what mushes? Face too small, text five words long, background competing with the subject, low contrast between the subject and everything behind it. Example: "At phone size I can read the word but not the face. Face bigger, word shorter."
Layer 3, taste. Color, style, whether you like the crop. This layer is real, it matters for brand consistency, and it should be last. Example: "We have used yellow on the last four. Let us break the pattern here."
When creators send me notes in that order, I can usually deliver a genuinely different second version rather than a cosmetic tweak. When notes arrive as a jumble, I have to guess which comment outranks which, and guessing costs a round.
The best feedback I ever get is one sentence long: "the promise is wrong, the video is actually about X". That single note is worth more than a paragraph about the drop shadow.
What should you say instead of "make it pop"?
Say what you want the viewer to feel or understand faster. "Make it pop" is a real instinct with a bad translation. It almost always means one of five specific things, and naming the right one gets you a fix immediately.
- "It does not stand out in the feed" means contrast or subject separation.
- "It looks flat" means the subject has no depth, lighting, or edge treatment against the background.
- "It looks cheap" usually means too many elements, or fonts fighting each other.
- "It is not exciting" means the promise is weak, not that the design is weak.
- "Something feels off" means the face expression does not match the title tone.
That last one is the sneakiest. A curious title with a shocked face, or an angry title with a smiling face, reads as noise. Viewers cannot articulate it, they just do not click. If you feel the mismatch, say the two words that describe the tone you want. "Should look skeptical, not thrilled." That is an actionable note.
A few more translations worth writing on a sticky note:
- Instead of "try something different", say what you would change about the promise.
- Instead of "more energy", name the emotion.
- Instead of "the text is weird", say how many words you want and whether it should be a claim, a number, or a question.
- Instead of "can we see a few options", say which decision you are unsure about. Options are cheap when the question is clear and useless when it is not.
How many revision rounds should a thumbnail need?
One or two, if the brief was good. Three or more means the brief was thin, and the feedback is now doing work the brief should have done. That is why I care so much about the front end. A tight thumbnail brief template filled in before any design starts removes most of the revision cycle before it exists, because the promise, the audience, and the competing videos are already agreed.
We include unlimited revisions in our packaging work, and I mean it, but unlimited revisions are not a substitute for a clear first instruction. Unlimited revisions are insurance for the cases where the video itself surprises you in the edit. They should not be the plan.
The second structural fix is deciding who is allowed to give notes. If a video has three stakeholders, and all three send comments directly, the designer receives contradictions and has to referee. Pick one person to consolidate. A simple thumbnail approval workflow with one named decision maker per video kills more revision rounds than any design skill I could teach you.
What does good feedback actually look like?
Here is the format I hand people. Copy it, keep it short, send it as plain text.
Promise: the video is really about [one sentence]. The thumbnail currently says [one sentence]. Closer to the first, please.
At small size: I can read [x]. I cannot read [y]. Priority is [y].
Tone: the face should read [emotion]. Right now it reads [emotion].
Taste: avoid [color or element] because [reason]. Optional.
Keep: the [element] is working, do not lose it.
That last line is the one people skip and it is the most valuable. If you never say what is working, every new version is a fresh gamble and the designer may quietly delete the best thing in the frame. Tell me what to protect.
One more habit: give feedback against data where you have it. If a similar video underperformed with a text heavy thumbnail, that is a fact, not a preference, and it settles arguments instantly. Our YouTube thumbnail analytics guide walks through which numbers in Studio are worth trusting and which are just noise from a small sample. YouTube's own creator resources are also decent on how impressions and click through rate relate, which is useful context before you start blaming a design for a distribution problem.
Does the feedback change for different formats?
Yes, in one important way. Long form thumbnails sell a topic. Podcast and interview thumbnails sell a person and a tension between two people. So the notes you give change shape.
On a podcast episode, the useful feedback is almost always about hierarchy: whose face is bigger, whose name matters to this audience, and what the single line of text claims about the conversation. Comments about background texture are close to irrelevant. I wrote about that in more depth in this piece on packaging for podcast creators, and it is the same principle we apply when the sister studio edits video podcasts.
On a new channel with no track record, feedback should lean much harder on clarity than on style, because you have no brand recognition to trade on yet. The notes should be ruthless about whether a stranger understands the topic in one glance. That is the whole argument in our guide to packaging a brand new channel.
Where should you practice this first?
On your old videos. This is the cheapest growth most channels are sitting on and almost nobody uses it. You already have videos with watch time and no impressions. Repackaging six of them is usually a better use of a month than shooting another video that lands in the same dead zone.
Old videos are also the perfect place to learn feedback, because the stakes are low and the data already exists. You know what the video is really about. You know what the audience did. You can write a sharp promise note without guessing.
My suggestion: take your five highest retention, lowest impression videos. Write the three layer note for each as if a designer were about to redo them. If you cannot write layer one in a single clear sentence, the problem was never the thumbnail.
Working with us on this
Our core package is $250 for 8 custom thumbnails, about $31 each, one per video, with unlimited revisions and a dedicated designer. Title consultation is included, because a thumbnail note and a title note are the same note most of the time. Turnaround is 48 to 72 hours per thumbnail. Add ons are simple: an A/B variant for $15, a title pack for $25, rush for $20.
Everything is custom per video. We do not run templates, and the styles we use get tested on our own channels before they touch a client's. We are based in Dubai and work worldwide and async, which is exactly why the feedback format above exists. When notes are clear, timezone gaps stop mattering.
If you want to see how we structure the work, the pricing and services page lays it out plainly. Send me a video you think was packaged badly and I will tell you honestly whether the design or the promise was the problem.