16:51Aprilynne Alter · Essay
Why No One Makes "Original" Videos Anymore
A YouTube strategist admits that the framework she teaches for chasing overperforming videos is quietly making every channel sound the same.
July 25thThe argument in one line.
As YouTube's interface shrinks titles and enlarges thumbnails across TV, mobile, and desktop, the thumbnail alone has to generate a question or an emotion, and the fastest way to design one is by reverse-engineering the psychological mechanism behind 10-15 gathered references, not by copying their imagery.
Read if. Skip if.
READ IF YOU ARE…
- You upload to YouTube regularly and design your own thumbnails without a dedicated designer.
- At least some meaningful share of your views come from TV, and you've never checked that number in Analytics.
- You tend to start a thumbnail by picking relevant imagery instead of asking what question or feeling it should create.
- You've never built a habit of collecting reference thumbnails before starting a new design.
SKIP IF…
- You outsource thumbnail design entirely to someone who already runs a reference/testing process.
- You don't publish on YouTube and have no thumbnail to design.
The full version, fast.
TV has passed mobile as YouTube's top viewing device, and TV's interface hides the bottom half of a thumbnail until a viewer presses down, so anything essential can't live below the halfway line. At the same time, every interface is shrinking titles relative to thumbnails, which means the thumbnail now has to generate interest on its own by provoking a question or an emotion, not just illustrating the topic. The reliable way to design one is to gather 10-15 references from five sources (your own outlier videos, other channels' outlier videos, format-matched channels in other niches, a bookmarked inspo playlist, and off-platform media), then ask what question or feeling each reference creates and how it does it, and apply that underlying mechanism to a new thumbnail rather than copying its look. Three fast fixes catch most of what's left: strip clutter, make the main subject visually obvious, and A/B test the thumbnail text.
Free for members
Chat with this breakdown — free.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
Create a free account →Where the time goes.
00:00 – 00:24
01 · Cold open
Five years of reviewing thumbnails distilled into the same recurring feedback; promises the most common notes in the next 10 minutes.
00:24 – 02:46
02 · Make your thumbnail TV-friendly
TV has overtaken mobile as the top YouTube device. TV's interface hides the bottom half of a thumbnail until a viewer presses down. Introduces the 50% test.
02:46 – 03:52
03 · Stop relying on your title
Across interfaces, thumbnails are growing and titles are shrinking, so the thumbnail alone must generate interest via questions or emotions. Speed's outfit thumbnail A/B compared.
03:52 – 05:21
04 · Stop starting from scratch: five sources of references
Gather 10-15 references from internal outliers, external outliers, Model 3 channels, a bookmarks playlist, and off-platform media (billboards, movies, Frameset).
05:21 – 06:40
05 · Extract the underlying mechanism
For every reference, ask what question/emotion it produces and how. Worked example: a 2D-vs-3D TV-intro frame reveals the 'subversion of expectations' mechanism, applied to make a 2D thumbnail look 3D.
06:40 – 07:43
06 · Sponsor: 1of10
1of10 used to visualize the 2D-to-3D thumbnail concept against the creator's own channel; $1 trial offer.
07:43 – 08:55
07 · Three quick fixes
Remove clutter, make the main character obvious (color + text label), A/B test the thumbnail text. Pool photo and Olympic diver worked examples.
08:55 – 09:55
08 · Recap + 90 Day Breakthrough CTA
Restates all six notes, pitches the 12-week 90 Day Breakthrough coaching program and the free thumbnail cheat sheet.
Lines worth screenshotting.
- TV has overtaken mobile as the primary device people use to watch YouTube in the US.
- YouTube's TV interface only shows the top half of a thumbnail until a viewer presses down on the remote, so anything below that line is invisible to a huge share of potential viewers.
- The 50% test: cover the bottom half of your thumbnail and check whether there's still enough there to make someone press down and to signal what the video is about.
- Across YouTube's interfaces, thumbnails are getting bigger and titles are getting smaller, so viewers now browse thumbnails alone and only read titles as a deliberate choice.
- A thumbnail now has to generate interest by itself; making it merely relevant to the title is no longer enough.
- The strongest thumbnails evoke a question, an emotion, or both, rather than just showing what the video is literally about.
- Start the thumbnail brainstorm by asking what you want the viewer to wonder or feel, not by asking what to show.
- Gather 10-15 references from five sources: your own overperforming videos, other channels' overperforming videos, format-matched channels in different niches, a saved bookmarks/inspo playlist, and off-platform sources like billboards, movies, and ads.
- When collecting references, don't worry about matching your exact topic; you're collecting approaches, not images to copy.
- For every reference, ask three questions: what question does it raise, what emotion does it evoke, and what specific mechanism produces that reaction.
- A 3D-looking image where 2D is expected creates novelty through the subversion of expectations, and that mechanism can transfer to an unrelated thumbnail concept.
- Removing distracting background clutter lets the eye land on the one element that matters in a thumbnail.
- The main subject of a thumbnail should be the first thing a viewer's eye lands on; color and text labeling can force that.
- A/B testing thumbnail text costs about five extra minutes and is the lowest-effort improvement available.
Your thumbnail has to work alone now, so stop copying and start extracting mechanisms.
WHAT TO LEARN
As titles shrink and TV viewing grows, a thumbnail has to generate a question or emotion on its own, and the reliable way to design one is to extract the psychological mechanism from a pool of gathered references rather than copy their imagery.
02Make your thumbnail TV-friendly
- Check your own TV viewership in YouTube Analytics under Audience > Device Type before assuming this note doesn't apply to you.
- Run the 50% test on your last thumbnail: cover the bottom half and check whether it still gives a reason to press down and a sense of the topic.
03Stop relying on your title
- Across YouTube's interfaces, thumbnails keep growing and titles keep shrinking, so treat the thumbnail as the primary sales pitch, not the title.
- Before choosing imagery, decide what question or emotion you want the thumbnail to create; let that decision drive the visual choices, not the other way around.
04Stop starting from scratch: five sources of references
- Build a standing habit of collecting 10-15 references from five sources: your own outlier videos, other channels' outlier videos, format-matched channels in other niches, a bookmarks/inspo playlist, and off-platform media.
- When you save a reference, don't worry about topic match; you're collecting approaches, not images to reuse.
05Extract the underlying mechanism
- For each reference, name the specific mechanism producing its question or emotion (e.g. subverted expectation, credibility signal) before trying to apply it to your own concept.
07Three quick fixes
- Strip any element from your thumbnail that isn't actively pulling attention toward the point; unrelated background detail works against you.
- Make the main subject the first thing the eye lands on, using color contrast or a short text label if needed.
- Spend five extra minutes generating alternate thumbnail text options and A/B test them; it is the lowest-effort lever available.
Terms worth knowing.
- 50% test
- Covering the bottom half of a thumbnail to check whether it still gives a TV viewer enough reason to press down on the remote and enough context to know what the video is about.
- Internal outliers
- A creator's own past videos that performed well above their channel's average, mined for thumbnail formats that might repeat.
- External outliers
- Other channels' videos that overperformed their own channel's average, found manually or with an outlier-search tool.
- Model 3 channels
- Channels that make a similar format of video to yours but cover a different subject, useful for spotting thumbnail formats that transfer across niches.
- Underlying mechanism
- The specific psychological trick a reference thumbnail uses to create a question or emotion (e.g. subverting an expectation), as distinct from its literal subject matter.
Things they pointed at.
Lines you could clip.
00:05
“90% of the time, I give the exact same feedback.”
02:08
“There is no reason to scroll down to get the full picture.”
02:38
“Titles are shrinking in relation to thumbnails.”
06:36
“When it comes to getting clicks, questions and emotions trump relevance every time.”
03:52
“The name of the game is numerous and varied.”
09:43
“Even tiny tweaks can make a huge difference.”
Word for word.
Read-along
Don't just watch it. Burn it in.
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
metaphor
Over the last five years, I've reviewed thousands of thumbnails from hundreds of creators, ranging from folks who are just starting out to those who've been at this for years. And recently, I've come to realize that 90 % of the time, I give the exact same feedback. So in the next 10 minutes, I want to walk you through the most common notes I keep giving.
Because of the creators I work with keep needing to hear them, odds are you'll find them helpful too. We'll start with one that lately I've been giving out more and more frequently, which is to make your thumbnail TV friendly. You see, in recent years, YouTube TV viewership has exploded.
In the US, TV has actually surpassed mobile as a primary device people watch YouTube on. To see what your own TV viewership is, head into analytics, click audience, and scroll down to device type. If at least 10 % is coming from TV, this note applies to you.
Because the current TV interface does something weird to thumbnails. In both the homepage and suggested views, only the top portion of recommended videos are visible. To see the full title and thumbnail, a viewer needs to press down on their remote.
But if all the most important elements of your thumbnail are hidden in the bottom half of your thumbnail, to a huge chunk of your potential viewers, your thumbnail looks like nothing at all. There is no reason to scroll down to get the full picture. So to make sure your TV viewers actually see the title thumbnail that you worked so hard to make, run your thumbnail through the 50 % test.
If you cover the bottom half of your thumbnail, is there still enough visible to prompt someone to press down on their remote? Is there enough there to give them a general idea of what your video is about? If the answer is no, it can usually be resolved by either moving your most important stuff to the top or simply adding text to the top third of your thumbnail.
But merely making your video's topic obvious isn't enough. Because if we take the TV view and compare it alongside other YouTube interfaces, one pattern makes itself abundantly clear. Across the board, titles are shrinking in relation to thumbnails.
Just take a look at the old versus new suggested panel. In the old version, while we did tend to glance at thumbnails first, the titles were large enough to be skimmed at the same time. Compare that to the new version.
The thumbnails are massive, and the title font size has actually gotten smaller. As a result, our eyes default to browsing the thumbnails alone, only reading a title as an active, intentional choice. Which leads to the second note I keep giving over and over again.
Stop relying on your title to generate interest. Before, as long as your title was interesting enough, all you had to do was make your thumbnail relevant and you'd be fine. But now, with your thumbnails receiving a disproportionate amount of attention, they need to be able to generate interest all on their own.
And the best way you can do that is by evoking questions, emotions, or ideally, both. Take this video from men's culture creator Speed. Now, most people would make the thumbnail by selecting some relevant imagery of outfits and accessories and call it a day.
But look at what Speed did instead. Taking away the titles, which thumbnail generates more interest? The second one.
Why? Well, it prompts us to wonder what are the five timeless rules and what will the final outfits look like? Meanwhile, the editorial font and lighting evokes a sense of credibility and taste, and seeing a grown man in his undergarments creates a subtle feeling of vulnerability.
And while the other thumbnail does show what the video is about, when it comes to getting clicks, questions and emotions trump relevance every time. So instead of starting your thumbnail brainstorming process with what should I show, consider first what you want your viewer to wonder and to feel.
Those questions and emotions will serve as a much stronger starting point to ideate from, ensuring that every element you come up with fulfills a specific psychological purpose. And if this kind of psychology -first brainstorming doesn't come naturally to you, that's okay. It's not expected to.
And in fact, that's exactly why the third note I keep repeating over and over again is stop starting from scratch. It continues to boggle my mind just how often someone will show me their thumbnail, I'll ask for inspiration, and they won't have any. And even if they do have references, it's rare that they actually use them in the right way.
But if there's one skill that will have by far the biggest impact on the clickability of your thumbnails, it's the ability to source and use inspiration correctly. So listen up. When it comes to thumbnail references, .
You're looking for between 10 to 15 different references from a variety of different sources. And there are five sources in particular that I found healed the best results. First, there's internal outliers, or past videos of your own that have overperformed your channel's average.
Could any of those thumbnail formats apply to your new video concept? Second, there's external outliers, or relevant videos from other people's videos that have overperformed their channel's average. You can hunt for these manually on YouTube, but I prefer using an outlier tool like 1 of 10.
which I'll link in the description. Third, there's Model 3 channels, or channels that make a similar format of videos to you, but who cover a different subject matter. Look at their top performing videos for any thumbnail formats that might translate to your video.
Fourth, there's bookmarks, or thumbnails that have caught your eye while browsing YouTube. If you don't have a system for saving these yet, a private thumbnail inspo playlist works great. And finally, there's off -platform inspiration.
Look at billboards, movies and TV shows, album covers, even ads you've been served recently. Frameset is a great free resource for this as well, which I'll also link below. As you gather your references, don't worry too much about matching your exact video topic.
All you're doing is collecting different thumbnail approaches. And once you have those approaches, you can move on to identifying their underlying mechanisms. For every reference you've pulled, ask, what questions, if any, does this make me ask?
What emotions, if any, does this make me feel? And most importantly, how does this reference make me ask those questions or feel those emotions? For example, when I was sourcing inspiration for the thumbnail of the video you're watching right now, I came across this frame from an intro of a TV show that I watch.
It didn't prompt any specific questions for me, but it did evoke a strong feeling. Something about it felt novel, like something I had never seen before. Looking closer, and with some help from Claude, which can be very useful for this, I realized the mechanism driving this novelty was the subversion of expectations by taking elements that are traditionally 2D and making them 3D.
Applying this mechanism to my own video concept, I knew that YouTube thumbnails are traditionally 2D, so what if I made one look 3D instead? This transferal step can be tricky, especially when your references are coming from off -platform or from somewhere outside of your own niche.
So to help visualize what this could look like, I used 1of10, who are also very kindly sponsoring this video. I gave 1of10 my reference and some thoughts about how it might apply to my thumbnail. And because 1of10 can link directly to your YouTube channel, using your own past videos as a style reference, it was able to show me a clear picture of what it could look like.
That picture gave me enough confidence to move forward with that thumbnail direction, and it served as the foundation for one of the thumbnails we're A -B testing on this very video. One of Ten's whole toolkit is incredibly useful, from the outlier search you saw earlier to ideation help, but their title thumbnail generators are truly the best of their kind I've ever seen.
If you want to check them out, I have a special link for you, which will let you try them out for a whole month for only $1, which I'll put in the video description below. Thank you, one of 10. Now, I know that sourcing and using inspiration like this can be a big process change.
The more you do it, the faster you'll get. But if you're looking for something that can improve your thumbnails right now, I got you.
Because the vast majority of thumbnails I look at can be dramatically improved with just three tiny changes. First, remove clutter. It doesn't matter what something looks like in real life.
Anything in your thumbnail that isn't actively working towards you is working against you. Take this thumbnail kindly submitted by your viewer.
This might be what the swimming pool actually looks like, but all of the bars are distracting for the most important parts of the thumbnail. Remove them and our eyes can much more easily focus on what they're supposed to be looking at. Second, make your main character obvious.
The main character of your thumbnail should be the first thing your viewers take in when they come across your video. So to make it easier for that to happen, make your main character stand out from everything else around it.
In this case, the Olympic diver is not obvious. Let's change the color of her swimsuit and add a text label. Now we instantly notice her.
And third, A -B test your thumbnail text. This is truly the lowest effort way you can improve the performance of your thumbnails. Whenever you add any kind of thumbnail text, just spend five extra minutes coming up with different words or phrases you could use instead.
I know these changes may seem small, but they're worth doing because when it comes to thumbnails, even tiny tweaks can make a huge difference. So if you're working on a thumbnail right now and have no one else to show it to, just take these notes from me.
Make it TV friendly, evoke more questions and or emotions, use more references, and reduce clutter, make the main character obvious, and come up with more text variations. And for more personalized feedback on your thumbnails, applications to 90 Day Breakthrough are open right now. 90 Day Breakthrough is my 12 -week hands -on YouTube accelerator program, where you'll work closely with me and my team to level up your channel through weekly assignments, personalized feedback on your videos, live sessions with content strategists, and much, much more.
The link to learn more and apply in the video description below. The description is also where you'll find the special link to try one of 10 for just $1, as well as a cheat sheet I put together for you to remember all of the notes covered in this video. So check those out below, but otherwise, thank you so much for watching.
Like and subscribe if you can. Keep creating and I'll see you in the next video.
The Hook
The bait, then the rug-pull.
Aprilynne Alter has reviewed thousands of thumbnails as a coach, and she's noticed that 90% of the time she's repeating the same handful of notes. This breakdown collects all six.
Named ideas worth stealing.
01:18concept
The 50% Test
Cover the bottom half of a thumbnail. If there's not enough left to make a TV viewer press down, or to signal the topic, the thumbnail fails on TV.
Steal forany thumbnail or cover image meant to survive a crop or a partial-visibility interface
04:22list
Five sources of thumbnail references
- Internal outliers
- External outliers
- Model 3 channels
- Bookmarked inspo playlist
- Off-platform media
A repeatable system for gathering 10-15 references before designing a new thumbnail, instead of starting from a blank page.
Steal forany visual design process (thumbnails, ads, landing pages) that currently starts from scratch
05:31concept
Mechanism extraction
For each reference, ask what question it raises, what emotion it evokes, and specifically how it does that. Apply the mechanism, not the image, to the new design.
Steal forswiping any creative reference (thumbnails, ad creative, hooks) without literally copying it
07:43list
Three quick fixes
- Remove clutter
- Make the main character obvious
- A/B test the thumbnail text
Fast, low-effort changes that improve most existing thumbnails without a full redesign.
Steal forany low-effort creative QA pass
How they asked for the click.
VERBAL ASK
09:20product
“For more personalized feedback on your thumbnails, applications to 90 Day Breakthrough are open right now.”
Soft recap of the video's own value first, then pivots straight into the coaching program pitch with program details (12 weeks, live sessions, personalized feedback), followed by a second CTA for the free thumbnail cheat sheet linked in the description.
MENTIONED ON CAMERA
FROM THE DESCRIPTION
PRIMARY CTAWhere the creator wants you to go next.
AFFILIATECommission earned if you click.
OTHER LINKSAlso linked in the description.
Visual structure at a glance.
