← All toolsRESEARCH YOU CAN CHECK

Understand the signal.
Inspect the evidence.

CreatorTube helps you find questions worth investigating. Every comparison depends on the videos collected and when their counts were observed.

Outlier score

Video views divided by the median of up to 20 earlier collected uploads of the same verified format, multiplied by a 1x-2x subscriber-reach bonus. At least five earlier uploads are required. The combined score is a ranking signal, not a views multiplier.

The default score compares lifetime counts. Older uploads can have an advantage. If the channel baseline is unavailable, including for unknown formats, the provisional score is views divided by a positive subscriber count, without an extra bonus. Its basis differs from the channel-based score.

Similar-age comparison

A separate snapshot ratio using same-channel, same-format peers with half to twice the target’s age when observed. At least five peers are required, with observations taken within seven days of one another.

This is a screening heuristic. It does not reconstruct how each video performed at exactly the same age.

Observed view growth

The difference between two dated public view-count observations, with the actual interval shown. Refreshing a video can add an observation; recent observations are reused for five minutes.

No points are invented between scans. Counts can decrease after corrections. Two observations are not a forecast.

Title patterns

The default comparison measures the share of sampled uploads matching five title families across two periods. Source videos remain available for inspection.

Upload share is not audience demand or view growth. AI title-structure interpretations are separate and require review; live AI output is not verified in this preview.

Channel comparisons

Median sampled video views and public subscriber counts help compare examples. Filters can change which videos contribute to the sample. Channel research offers 30- and 90-day publication windows; their view counts are lifetime observations, not views earned in those windows.

This is not a complete channel catalog. Country and language are creator-reported metadata, not audience location. Revenue, RPM and monetization are not verified.

Lifetime views per hour

Observed views divided by the video’s age at that observation, with a one-hour minimum denominator. Missing counts or invalid dates remain unavailable.

This is a lifetime average, not current hourly velocity. Use dated observations to inspect actual change.

Search positions

A dated result from the YouTube Data API for a keyword and country. Later checks build a history of observations.

Personalized YouTube search can differ. Missing checks stay missing; enabled schedules run through hosted background alarms or the running local development server, subject to access and resource limits.

A shortlist, not a guarantee.

Open the references, check the context and develop your own angle. The preview uses a limited imported catalog, not all of YouTube.

Explore video research →