See the evidence.
Understand the limits.

Observed data and modelled estimates

Public channel and video metadata, subscriber counts and view counts come from the YouTube Data API. These are indexed snapshots, not live readings of every channel. BlipScope does not know another creator’s private revenue, retention or audience analytics.

NICHE VALIDATOR

From your question to a comparison

1. Find relevant indexed channels

Your niche description and optional audience text form one semantic query. The production search considers up to 100 nearest candidates per search and returns up to 60 eligible comparison channels. Eligibility requires at least 1,000 subscribers and a positive monthly-view estimate. This is our comparison rule; it does not verify monetization or exclude every inactive channel.

2. Check whether the index supports an answer

A scored result needs at least 8 comparable channels, a minimum text similarity of 0.22 and enough separation from the background index. The separation check compares the mean similarity of the 10 nearest channels with a conservative median estimate from a random background sample, requiring a gap of at least 0.135.

These are product safeguards, not a statistical guarantee. The matching channels, contributing sample counts and available observation dates appear with the result. The background sample tests relevance; it is not the sample used to calculate the niche’s RPM band.

3. Show a score, or acknowledge insufficient coverage

The score combines estimated ad rates, modelled commercial assumptions, subscriber concentration and available outlier scores. Weights change with your selected monetization plan. Fewer than 5 available outlier scores means that component sits out and the remaining weights are rescaled. Revenue and faceless cards provide context; they are not separate weighted inputs to the overall score.

A score of 72 is not a 72% chance of success. If the sample or relevance checks fail, the validator returns no score. Limited coverage can reflect the index or wording; it does not establish an untapped market.

RPM & REVENUE

A range of estimates is still an estimate.

RPM starts from assumptions

The model starts with a category ad-rate assumption and adjusts for the channel’s declared country, subscriber scale, recent long-form video duration and a made-for-kids classification. The declared country is a proxy: it does not tell us where the viewers actually live. Public data does not establish whether a channel is monetized.

Estimated monthly ad revenue combines modelled RPM with estimated monthly views and a Shorts adjustment. Where usable observation history exists, monthly views are extrapolated from changes in channel views; otherwise a public-data estimate is used. Sponsorships, affiliate sales, products, memberships, taxes and production costs are not included.

What the band tells you

The RPM band is the 25th–75th percentile: the middle half of the comparison channels’ estimates. Monthly revenue shows the median and 90th percentile of estimates. These describe variation within the selected sample, not uncertainty around a verified earnings figure. They are not confidence intervals or forecasts for a new channel.

Extreme values are retained. The validator does not remove statistical outliers or let you trim them. Percentiles reduce the influence of extremes compared with a mean, but a small or biased sample can still mislead. Inspect the channel list and refine the query when matches do not fit.

Money estimates are displayed in USD for comparison. Plan prices are billed in EUR. BlipScope does not convert estimated earnings into your billing currency.

Outlier scores and faceless flags

A channel outlier score is not a video multiplier

The channel score runs from 0 to 6. It compares median recent-video views per subscriber with the channel’s category baseline. Each point represents a doubling above that baseline: 1 means 2×, 2 means 4× and 3 means 8×. Zero includes channels at or below the baseline. Values are capped at 6.

Only scores marked available contribute. A valid zero counts; an unknown, pending or otherwise unavailable score does not. This comparison is not adjusted for upload age or video format and does not prove that YouTube is promoting a channel. Video-level multipliers shown elsewhere use different reference uploads.

“Faceless” is an automated classification

Metadata rules and available AI enrichment flag likely faceless channels. They do not amount to manually watching every upload. A flag can be wrong or stale, and no flag does not prove an on-camera presenter. Watch several recent uploads before deciding whether the channel matches your intended format.

Commercial potential is a category assumption

Affiliate and offer scores use category assumptions weighted by the sample’s category mix. They do not measure actual affiliate links, sales, conversion rates or willingness to pay among those channels’ viewers.

Freshness, coverage and your next step

The validator shows the oldest and newest available public-data observation dates in its sample, in UTC. A missing date is disclosed as unknown. Channel detail pages let you inspect individual records. Refresh schedules and queues mean indexed counts can differ from YouTube; the page does not promise a live refresh.

The index is a selected set of channels. Language, geography, discovery queries and the comparison rules affect who appears. At least 8 matches is a minimum guard, not proof of a representative market sample. Strong similarity is text relevance, not measured audience overlap.

Keep YouTube Studio in your workflow. Use your own Studio analytics to learn how your published videos perform. Use BlipScope to investigate public examples and compare ideas before production. Check source channels, make a small original test and evaluate the outcome in your own analytics.