What is a YouTube outlier, and how to use one to pick your next topic

Scroll through any niche on YouTube and the biggest numbers jump out first: the 4 million view video, the channel with 800,000 subscribers. They are also the least useful numbers on the page. A big channel getting big views tells you almost nothing about whether you could do the same, because most of those views come from an audience the channel spent years building.

An outlier is a much sharper signal. It is a video, or a channel, that performs far above what you would expect from it. The word "expect" is doing the heavy lifting, so let's be precise about it.

The definition: views compared to a baseline

An outlier score is a ratio. You take a result and divide it by a baseline that describes "normal" for that situation.

At ByNiche we use two versions:

• Video outlier = the video's views ÷ the channel's median views.
• Channel outlier = the channel's average views per video ÷ the median of comparable channels in the same niche and format.

So a video with an outlier score of 8× pulled eight times the views its channel usually gets. A channel with a score of 5× averages five times more views per video than a typical channel making the same kind of content.

That framing removes the main thing that distorts raw view counts: the size of the audience that was already there.

Why the median, not the average

Averages are easily pulled around by a single viral upload. Imagine a channel whose videos usually land somewhere around 3,000 views, plus one that reached 600,000. Add those up and the average suddenly looks like a channel that does 40,000 views per video. It doesn't. Almost none of its videos come close.

The median is the middle value when you sort the videos by views. One runaway hit barely moves it. That makes it a far more honest picture of what a "normal" upload looks like on that channel, which is exactly what you want to compare against.

Two more details matter if you want the baseline to be fair:

1. Keep formats apart. Shorts and long-form videos live in different worlds. A Short with 200,000 views and a 20-minute documentary with 200,000 views are not the same achievement. ByNiche measures each channel in its dominant format and compares it only with channels in the same format.
2. Ignore videos that are too new. A video uploaded yesterday hasn't had time to collect its views yet, and including it drags the median down. We leave out videos younger than three days when there are enough older ones to work with.

Why the niche median matters for channels

The video outlier answers "what worked unusually well for this channel?" The channel outlier answers a different question: "is this channel doing unusually well for its niche?"

Niches have very different natural levels. A typical small channel explaining tax rules and a typical small channel posting gaming clips do not get the same views, and that's fine. Comparing a channel to its own niche tells you whether something special is going on, regardless of where the niche sits overall.

When a niche doesn't have enough tracked channels to produce a stable median, ByNiche falls back to the broader category in the same format, and then to the format alone. That keeps the comparison meaningful instead of comparing against a handful of channels.

What outliers tell you that raw views can't

Here is the core idea: when a small channel gets an outlier, the topic did the work. The channel didn't have a huge subscriber base to push the video. The thumbnail and title were probably no better than the channel's other uploads. Something about the subject itself made YouTube show it to many more people, and made those people click.

That is the most transferable thing on YouTube. You can't borrow someone's audience, but you can study what made a topic travel.

A large channel's popular video, by contrast, blends the topic with a loyal audience, notifications and years of recommendations. You can't tell which part the topic earned.

How to use outliers to pick topics

A practical routine looks like this.

1. Collect outliers from channels your size or smaller

Filter for channels that are still small, then look at their videos with the highest outlier scores. A 10× video from a channel with 4,000 subscribers is worth much more of your attention than a 2× video from a channel with 2 million.

2. Look for the pattern, not the single video

One outlier can be luck. Three outliers on related topics across different channels is a pattern. When you see several small channels in a niche breaking out on, say, "the history of a specific product" or "explaining a confusing everyday document", that's a theme the audience is hungry for.

Group what you find by:

• Subject: what is the video actually about?
• Angle: is it a story, a ranking, a mistake to avoid, a comparison?
• Title structure: a question, a number, a strong claim, a "why" or "how"?

3. Check whether the channel repeats it

A channel with one huge outlier and nothing else might simply have caught a lucky wave. A channel that breaks out again and again probably understands something about its audience. ByNiche's consistency label looks at the channel's recent uploads and tells you whether it's a single hit, has repeat hits, is steady (all videos land close together) or uneven. Repeat hits are the strongest evidence that a topic area is reliably in demand, not a one-off.

4. Make it yours

The goal is not to copy a video. It's to understand the demand behind it and answer it better or from a different angle. If an outlier explains a topic in ten minutes, maybe there's room for a deeper version, a version for beginners, or a version for a specific country or audience.

5. Watch the recency

An outlier from last month says more about today's audience than one from three years ago. Interests shift, and the recommendation system shifts with them. Prioritise recent breakouts, and treat old ones as background.

Common traps

Unusual engagement. Some channels get lots of views but almost no likes. That can have many causes, such as ad-driven or embedded plays, and it makes the view count a weak signal of real demand. ByNiche keeps channels whose typical like rate is unusually low on videos with substantial views out of the default results, so you don't build a plan on demand you can't verify.

Tiny samples. A channel with three videos doesn't have a meaningful median yet. Give more weight to outliers from channels that have uploaded enough to establish a baseline.

Chasing the number alone. A 30× outlier on a topic you have no interest in, and no way to produce well, isn't a good topic for you. Use outliers as a filter on top of what you can actually make.

A quick worked example

Say you're exploring a history niche. You notice a channel with a few thousand subscribers whose videos usually sit around a few thousand views. One of its recent uploads, about an everyday object's surprising origin, reached ten times that. Looking further, you find two other small history channels with similar "surprising origin" videos that also beat their channel medians by a wide margin.

That's three independent signals pointing at the same angle. You still need to check competition and whether you can produce that kind of video well, but you now have a topic direction backed by evidence instead of a guess.

(The numbers here are illustrative, not taken from a specific channel.)

The short version

• Raw views mostly measure audience size. Outliers measure what made people click.
• Compare a video to its channel's median, and a channel to its niche's median, within the same format.
• Small channels with repeat outliers are the best evidence of real demand.
• Use outliers to find topics and angles, then make your own version.

ByNiche rescans thousands of channels every day and calculates these scores for you, so you can go straight to the interesting part: deciding what to make next.