Skip to content
channelpeeker

Methodology

Last updated November 7, 2026

We want every number on channelpeeker to be checkable. This page explains where our data comes from, which figures are ours, and how we calculate them.

Two kinds of numbers

  • YouTube data: public figures from the official YouTube Data API, such as view counts, like counts, comment counts, upload dates and subscriber counts.
  • channelpeeker estimates: anything we calculate, such as medians, ratios, outlier multipliers, scores and earnings ranges. These are always labeled as channelpeeker estimates. They are not provided or approved by YouTube or Google.

Known limits of public data

  • Subscriber counts are rounded. YouTube rounds public subscriber counts to three significant figures above 1,000 (123,456 shows as 123,000). We show them with a ≈ sign. View counts are exact.
  • Dislikes are private. We don't show like/dislike ratios.
  • Watch hours, CTR and revenue are private. Only the channel owner can see them in YouTube Studio.
  • View counting changed in 2026. YouTube announced that public views count from the first frame for all formats. Charts that span that change may show a step.

Tools we deliberately don't build

Some popular tool ideas can't be answered honestly from public data, so you won't find them here even though competitors offer them:

  • Shadowban checkers. YouTube has no public "shadowban" flag. Sites offering this guess from search results, which produces confident-looking answers nobody can verify.
  • Dislike checkers. YouTube removed dislike counts from its API in 2021. Any site showing current dislikes is estimating, not reporting.
  • Deleted or private video viewers. The official API doesn't return deleted or private content. Tools claiming to recover it are scraping third-party archives of varying legality and accuracy.
  • Monetization verdicts. Since 2023 YouTube exposes no "is monetized" flag. Our Monetization Checker shows which Partner Program requirements a channel visibly meets instead of inventing a yes/no.

If a number here ever disagrees with a site that does offer those tools, the difference is usually that we won't show data YouTube doesn't publish.

Definitions

Median views

The middle value of a set of videos' view counts. We prefer medians to averages because one viral video can distort an average.

Outlier multiplier

A video's views divided by the median views of the same channel's recent videos of the same format (Shorts or long-form). A 5× outlier got five times the typical views.

Engagement ratios

Likes ÷ views and comments ÷ views, shown as percentages.

Earnings estimates

Earnings are shown as ranges, never as one exact number. Real earnings depend on factors no outside tool can see, including ad inventory, viewer location, watch time and advertiser demand. The formula behind our revenue calculators:

earnings = views ÷ 1,000 × monetized share (default 60%) × 55% creator split × niche CPM range × audience multiplier

CPM table (long-form, USD, advertiser CPM)

These are channelpeeker estimates of industry-typical ranges, reviewed quarterly. Last reviewed: November 2026. They are not YouTube data.

NicheCPM range
Finance & investing$15–$40
Business & marketing$10–$25
Tech & software$8–$20
Education & how-to$6–$15
Health & fitness$5–$12
Beauty & fashion$4–$10
Travel & outdoors$4–$10
Food & cooking$3–$8
Gaming$2–$6
Entertainment & vlogs$2–$5
Music$1–$4
Other / mixed$2–$8

Audience multipliers

Main audienceMultiplier
US, UK, Canada, Australia×1
Western Europe, Japan, Korea×0.7
Eastern Europe, Latin America, SE Asia×0.35
South Asia, Africa, other×0.15

Shorts are estimated separately at $0.04–$0.15 creator RPM per 1,000 views. Sponsorship rates use $20–$50 per 1,000 expected views for a dedicated integration.

Keyword data

Keyword tools use YouTube's public search suggestions — real phrases people type. We don't show monthly search volumes because Google doesn't publish them for YouTube; any precise volume number you see elsewhere is an estimate. Keyword Difficulty shows the top-10 ranking videos from the official API; because a search costs 100 quota units, results are cached for 7 days.

AI tools

Generator and summarizer tools send your topic or pasted transcript to Google's Gemini API to produce suggestions, and are limited to 10 generations per day. We don't store the inputs. AI output is a draft: review it before publishing. Transcript tools work on text you paste from YouTube's own transcript panel, because YouTube's API only allows caption downloads by a video's owner — we don't scrape private endpoints.

Freshness

We cache API responses to stay fast and within YouTube's quota. Every result shows when it was fetched. See the data policy for how long we keep each type of data.