TikTok Tests Your Video. YouTube Shorts Might Not.

I publish a lot of short-form video. Enough that I’ve stopped looking at every post as an isolated win or loss and started paying much more attention to how the platforms themselves behave.

One of the clearest differences I’ve experienced between TikTok and YouTube Shorts is surprisingly simple:

TikTok seems to test almost everything. YouTube sometimes just… doesn’t.

I don’t mean a YouTube Short gets a small test, performs badly, and dies. I mean it can receive virtually no meaningful Shorts Feed distribution at all.

I’ve seen that happen enough times that I now make an important distinction between a video that failed its test and a video that never really got tested.

Those are not the same thing.

TikTok usually gives the video a shot

My experience with TikTok has generally been straightforward. I post a video, TikTok gives it an initial audience, and the video starts proving — or failing to prove — that it deserves more distribution.

Sometimes the first test is small. Sometimes it expands quickly. Sometimes a video gets a few hundred views and dies. But it usually feels like the video got its shot.

That makes TikTok relatively easy to reason about.

If a post gets distribution but people swipe away immediately, retention is weak, and nobody engages, I can learn something from that. Maybe the opening didn’t work. Maybe the joke took too long. Maybe the premise wasn’t interesting enough. Maybe it was simply a bad video.

Either way, the platform showed it to people. People responded. Now I have data.

YouTube Shorts can be a completely different experience

There are periods where I upload a Short and YouTube immediately puts it into the Shorts Feed. On my channel, a normal first meaningful test during those periods has often landed somewhere around 1,500 to 2,700 views.

That’s enough traffic to evaluate the creative. I can see how often viewers chose to watch instead of swiping away. I can look at retention, average percentage viewed, likes, and the rest of the audience-response signals.

Then there are other periods where I upload a Short and essentially nothing happens.

Not 1,500 views. Not 500. Sometimes there is barely any Shorts Feed traffic at all.

And I’m not talking about one random upload.

During 2026, the pattern became especially noticeable on my channel. Beginning around April, I started seeing a much larger share of Shorts receive extremely low distribution or apparently fail to enter the Shorts Feed in any meaningful way.

Some of those stretches lasted days. Others lasted weeks. The longest lasted roughly a month.

Then distribution would start flowing again.

I informally started calling these periods “open” and “frozen”, but those are only descriptions of what I can see in my analytics. I have no evidence YouTube has an internal mechanism that works that way.

What I do have is a lot of uploads and a very noticeable difference in whether those uploads receive enough Shorts Feed exposure to produce useful audience data.

Then one of my “dead” Shorts came back to life

The strangest part is that some Shorts that initially appear dead eventually get tested.

One of the clearest examples on my channel was a video I internally called “Erling Googley Eyes.” At first, it looked like another Short YouTube simply wasn’t going to distribute.

Then it started moving.

Eventually, the Short reached roughly 18,000 to 20,000 organic views.

Once YouTube actually put it in front of people, the audience response was exceptionally strong. Its stayed-to-watch rate was around 74%, and its average percentage viewed climbed into roughly the 170% range, meaning viewers were frequently looping or rewatching it.

The creative hadn’t suddenly changed. It was the same video.

The meaningful change was that YouTube finally gave it an audience.

If I had judged that Short exclusively from its initial view count, I would have concluded that it was a failure.

It wasn’t. It just hadn’t been meaningfully tested yet.

Distribution is not the same thing as performance

Creators naturally look at views as the scoreboard, but I don’t think views alone tell you enough.

I’ve started thinking about short-form distribution as three separate stages:

Allocation → Audience response → Expansion

First, did the platform actually give the video an audience?

Second, how did that audience respond?

Third, did the platform expand distribution afterward?

Imagine two Shorts.

The first gets shown to 2,000 people. Most viewers immediately swipe away. Retention is terrible and nobody engages.

The second receives almost no Shorts Feed exposure at all.

Twenty-four hours later, both may look like failures. But they tell you completely different things.

The first video probably failed its test.

The second hasn’t told you much of anything.

This is also why I’m skeptical of the countless formulas floating around creator communities: “Hit this retention percentage and YouTube will push your Short,” or “Reach this viewed-versus-swiped threshold and you’ll get another wave.”

Those numbers can be useful benchmarks, but my own uploads have not behaved like a simple vending machine where you insert the right retention rate and receive a predictable number of views.

I’ve seen relatively mediocre audience-response numbers receive a respectable test while stronger videos stalled.

There are clearly other variables involved, and I don’t pretend to know exactly what YouTube’s recommendation system is doing behind the scenes.

I now treat some Shorts as “untested inventory”

This difference matters most because of what it does to the creative feedback loop.

TikTok gives me fast feedback. If nearly every video receives some kind of initial test, I’m constantly learning. Did people stop scrolling? Did they make it through the opening? Did they finish? Did they rewatch? Did they share it?

YouTube’s inconsistent distribution makes that harder.

If a Short receives almost no Shorts Feed exposure, there isn’t enough audience behavior to diagnose the creative. One of the easiest mistakes a creator can make in that situation is to start “fixing” a video based on data that never really existed.

So I changed the terminology I use for my own channel.

If a TikTok gets its normal test and viewers reject it, I’m comfortable calling it a miss. If a YouTube Short receives meaningful Shorts Feed distribution and the audience rejects it, same thing.

But if a YouTube Short gets virtually no Shorts Feed exposure, I don’t call it a failure anymore.

I call it untested inventory.

Maybe it’s great. Maybe it’s terrible. I don’t know yet. The audience hasn’t answered the question.

And after publishing hundreds of short-form videos, that may be the biggest difference I’ve experienced between TikTok and YouTube Shorts:

TikTok usually lets the audience decide whether my video deserves to live.

YouTube sometimes hasn’t asked them yet.