Outliers

An outlier video is a video that received significantly more views than the rest of the channel usually does. If a channel averages 10,000 views per video and then suddenly has one with 1,000,000 views, that is an outlier. The same can be said when a new channel has only a few thousand subs but is getting 50k, 100k, 300k+ on their first videos. YouTube pushed these videos because demand was there, but the platform's supply was missing.

We track these with Outlier Scores. Ours is the ZYRA Score, and it is the one to read first. It runs 0 to 100, and this is how rare each band is inside the Vault:

HYPER: 100+ · 1% of the Vault
ULTRA: 90 - 100 · 4.7%
MEGA: 80 - 90 · 12.7%
SUPER: 60 - 80 · 25.1%
NORMAL: 35 - 60 · 30.5%
CHALK: under 35 · 26.1%

Next to it sit the other two scores: how far the video beat the channel's average, and how far it beat what its subscriber count says it should get. Those two are plain multiples, and they read on this ladder:

Normal outliers: 2 - 4x
Super outliers: 5 - 10x
Mega outliers: 10 - 30x
Ultra outliers: 30x+

These outlier scores are easily trackable with extensions, or simply using the Metarch to log any interesting outliers you find while using the Outlier Vault. Every outlier tool out there bases their outlier scores on different metrics… some take the last 6 videos before an outlier video into account, some 30, and some even 100.

In 117's Metarch Outlier Vault, we give the most comprehensive stats possible for you to assess REAL outliers. We give you the average over the past 20 videos for recent outliers, we give the rolling 12 month average for older ones, and we include the sub outlier score as well. But the ZYRA Score is built on the MEDIAN, which no one else does. Why?

Because a channel with 10 videos and a million view video looks great when you average it out: 100k view average! But.. what if the rest of the videos are literally getting hundreds and low thousands? The average is an absolute meaningless statistic. The median gives you a better indicator of the TRUE performance of such a channel. But there is much nuance about which is best and when, and that nuance is exactly what the ZYRA Score does for you. It starts with the median, then weighs how wildly that channel's views swing, how often they post, and how big they are, and hands you ONE number out of 100. Here is why that matters: a 3x on a channel that swings all over the place is luck. A 3x on a channel that never misses is a real signal. Every other tool prints 3x for both. ZYRA does not. You still get the average and the sub score sitting right next to it, so you can always see the raw numbers it was built from.

Ultimately, the most important thing to understand about outliers: the most valuable ones are not one-off lucky hits. The strongest niches are the ones where outliers keep appearing across multiple channels, on formats you can realistically produce. That means: If it costs $10-20k per video to make? It's probably not worth the risk.

Authority Channels

Authority channels are the dominant channels in a niche. They typically have a large subscriber base (often 1 million+, 500k+, 250k+, though this varies by niche), a strong returning-viewer base, recognizable, repeatable formats, and strong audience trust built over time.

Authority channels tell you what the niche's creative standard looks like, which formats already have proven demand, what topics are evergreen, and what the audience already understands and expects. Viewer pyschology is fundamental to you being succesful. Put your "viewer hat" on, and put yourself in their position. Why do they love these authority channels?

There is a catch though: their videos can perform well partly because of their existing subscriber base, and not purely because of the format or topic. They are there for the personality as well. You use them to understand the niche's packaging, video style, and recurring topics. They are your guide of what already works.

Outlier Channels

Breakthrough or Outlier channels are smaller or newer channels pulling far more views than their size would suggest. They typically have relatively low subscriber counts compared to their views, and often use formats or angles that feel less saturated than those of the authority channels. However, the formats/angles can also be torched to death across man niches and it might be wise to consider avoiding them if you're entering the space too late... such as "Why It Sucks to be"

Breakthrough channels tell you where the niche is expanding right now, what new formats the algorithm (read as: the audience) is currently rewarding, and whether there is still room for a new channel to enter. When a small channel gets huge views, that is almost always the algorithm signaling that demand exists, but supply is weak. That gap is exactly what you are looking for. However, you have to keep tabs on how many Breatkrhough Channels are rising... use your Discovery tool to track these but if there are a lot popping up: How are the views looking? Rising, stable, or falling? The maturation of the market/niche is important to assess as well. Entering too late in formats that have been heavily used is often a negative action for your success.

Ultimately, Authority channels show you what is proven. Breakthrough channels show you what is emerging. Both matter for your niche research but outlier channels are where the clearest signals of opportunity come from. Just try to be earlier rathern than later.

Niche

Oh, right.. just what isa niche? Well, it's not just a topic, it's an economic market of attention! Simply, a niche is the specific audience + specific style + specific format your channel is built around. "History" is not technically a niche, but we use these macro terms when STARTING our research, of course. "World War II Military Strategy" is a more specific niche. "Middle Ages Politics" is also History but these are VASTLY different audiences.

The mistake most people make is picking a topic that interests them and calling it a niche. The result is a channel that has no consistent viewer, no consistent packaging, and no clear reason for the algorithm to push it to find an audience. A real niche has an obvious answer to: who is the returning viewer, and what do they get every time?

Your goal here isn't to lock in your niche RIGHT NOW, but to test three potential niches against real data (authority channels and outlier channels) until one starts looking like you might just have a powerful opportunity!