Blog

Notes on making videos and Shorts, based on real YouTube data.

  1. Manual Chapters Not Showing? Check the Timestamps First

    Check chapter intervals with a fictional two-minute video and verify the saved display.

  2. What Comes After the Ending? An End-Screen Handoff Worksheet

    Plan a useful next-video handoff and distinguish end-screen eligibility, display and performance checks.

  3. Views Rose but CTR Fell: Check the Denominator and Traffic Mix

    Use a hypothetical traffic mix to understand aggregate CTR, then complete a worksheet before changing your thumbnail.

  4. When Estimated and Finalized YouTube Revenue Differ: A Monthly Reconciliation

    Separate estimates, finalized earnings and bank receipts, then match dates, currency and documented adjustments.

  5. Why Averaging Video RPMs Can Mislead: Weighted RPM and Revenue Checks

    Calculate weighted RPM with fictional examples and reconcile Studio totals by matching dates, currency, denominators and revenue scope.

  6. A Publish-Time Experiment for Your Channel: Compare Equal 24-Hour and 7-Day Windows

    Track publication time, timezone and format, then compare equal elapsed windows without confusing small samples or content differences with timing effects.

  7. Thumbnail A/B Tests: Why the Highest CTR Is Not Automatically the Winner

    A practical worksheet for YouTube title and thumbnail tests: read watch-time share, interpret result states, and plan the next experiment.

  8. Shorts Views vs Engaged Views: Five Checks Before Calculating Revenue

    A practical Studio-to-calculator checklist for scope, dates, engaged views, RPM and currency, with an illustrative example of a public-view input mistake.

  9. Correction: Lengths of 658 Videos Up to Three Minutes and Sample Limits

    Observed lengths and topic differences in 658 videos up to three minutes. Shorts classification was not verified; this correction explains why the sample cannot establish optimal length or earnings.

  10. Top 4–20 Minute Search Results Ran About 12 Minutes; the Popular Chart Ran Longer and Was Mostly Gaming

    We compared the length in minutes of 619 top 4–20 minute search results with videos on the Korean, US and Japanese popular charts.

  11. Seven in Ten Top Short-Video Search Results Had More Views Than Their Channel Had Subscribers

    We checked subscriber counts of channels behind popular videos, split by top search results and the popular charts, to see where small channels show up.

  12. Three Days Later, 25 of 200 Videos Were Still on Korea's Popular Chart

    We compared the Korean, US and Japanese popular charts on October 23 and 26 to see how many videos changed and which ones stayed.

  13. Popular Video Descriptions: Brief on Short Videos, Chapters and Links on Long Ones

    We counted description length, chapters, links and hashtags across about 1,000 popular videos, split by videos of three minutes or less and longer ones, and by Korea, the US and Japan.

  14. Two in Three Popular Videos Under Three Minutes Had No Tags

    We checked whether popular videos used tags, how many, and how often tags repeated the title, split by videos of three minutes or less and longer ones, alongside what YouTube says tags do.

  15. Why Public Views Cannot Compare Shorts Earnings: October Sample Correction

    We correct a Shorts RPM comparison based on public views, retaining sample counts and medians from four topics and explaining engaged views and the limits of the data.

  16. Most Popular-Chart Videos Were About a Day Old: Korea, US and Japan

    How many hours old the 599 videos on the Korean, US and Japanese popular charts were, and what length and category they had.

  17. Comments and Likes by Duration: Search Sample Correction

    Preserved reaction rates per 1,000 public views, with corrected Shorts and long-video advice and explicit denominator, missing-data and causal limits.

  18. Popular Korean Videos Were Most Often Posted Between 6 and 9 p.m.

    We checked what local time popular videos and popular-chart videos were posted in Korea, the US and Japan.

  19. 4–20 Minute Video Titles: Search Sample Comparison and Correction

    Historical title observations from 140 videos of 4–20 minutes and 100 videos of three minutes or less, with different search times and corrected classification limits.

  20. Why YouTube's Popular Chart Shows Mostly Games and Music

    What 399 videos on the Korean and US popular charts looked like in October 2026, and what to check instead when looking for video ideas.

  21. Video Length by Topic: October Search Sample Correction

    Preserved duration counts from 260 search results across 13 topics, with a correction: length alone cannot establish Shorts classification or production priority.

  22. Video Title Patterns: Correction to a 260-Result Search Sample

    Preserved title observations by duration, with corrections to Shorts classification and title-length and hashtag advice. Raw-data verification limits are disclosed.