How to analyze a YouTube Analytics CSV and find videos that drive growth
Analyze YouTube performance at video and time-period grain, separating reach, packaging, viewing quality, and audience response. Compare impressions, click-through, watch time, retention, subscribers, traffic source, topic, and video age together; raw views alone reward older or heavily distributed videos.
Why the obvious method fails
A recent video has less lifetime views than an older one but may convert impressions and subscribers better. Shorts and long-form videos have different viewing behavior, and one viral result can dominate channel averages.
Growth comes from a sequence, not one metric
Impressions create opportunity, packaging earns a view, content holds attention, and the viewer may return or subscribe. Diagnosing the stage prevents a creator from changing video content when the actual constraint is distribution or thumbnail fit.
Where this usually goes wrong
Platform attribution and metric definitions can change. Record export dates and consult current platform documentation before treating fields as stable across long histories.
A useful mental model
- Compare the first 7, 28, or 90 days after publish as well as lifetime totals.
- Use impressions and CTR for packaging, retention and watch time for viewing, and subscribers or actions for outcome.
- Separate format, topic, length, traffic source, and publishing cadence.
- Use several videos and uncertainty, not one viral outlier, before changing strategy.
From the field · YouTube creator
Identify which videos attract viewers who stay and subscribe
The creator exports video-level and daily analytics. Tutorials receive fewer impressions than trend videos but higher watch time and subscriber gain; one trend video went viral through external traffic.
Columns used in the example
video_id · publish_date · format · topic · impressions · ctr · average_view_duration · watch_time · subscribers_gained · traffic_source
Follow it with your own file
- Join exports by video_id and create age-aligned performance windows.
- Separate Shorts, long-form, topic, length, and traffic source.
- Compare reach, packaging, viewing, and subscriber outcomes with denominators.
- Select patterns supported by several videos and propose the next content test.
Check: Metrics must use consistent windows and definitions, video counts must reconcile to exports, and the conclusion should be tested with and without the viral outlier.
Choose the approach that matches the evidence and repetition
| Approach | When it works | What to verify |
|---|---|---|
| Lifetime views | Overall historical reach | Video age and distribution |
| First-28-day funnel | Comparable launch performance | Season and source mix |
| Topic/format pattern | Content planning | Sample size and outliers |
Patterns that need context
- Videos are ranked only by lifetime views.
- Shorts and long-form retention are compared directly.
- Subscriber gains are not aligned with video publication and attribution windows.
A method you can follow
- Set comparable windows: Compare the first 7, 28, or 90 days after publish as well as lifetime totals.
- Separate the journey: Use impressions and CTR for packaging, retention and watch time for viewing, and subscribers or actions for outcome.
- Segment context: Separate format, topic, length, traffic source, and publishing cadence.
- Find repeatable patterns: Use several videos and uncertainty, not one viral outlier, before changing strategy.
Doing these steps in Prepyr
Prepyr accepts CSV files only. Export the worksheet or table you need as a UTF-8 CSV first; workbook formulas, formatting, macros, charts, and other sheets will not be included. Describe the result you need: Analyze a YouTube Analytics CSV to find repeatable video growth drivers. Check the columns selected, preview any proposed changes, and compare the output with the source before applying or exporting it.
Prepyr steps used here: Multi File analysis · Charts · Reports · Reusable workflows
What to verify before acting
- Video IDs aligned
- Performance windows comparable
- Formats separated
- Outliers tested
- Next experiment defined
A creator report that identifies repeatable combinations of topic, packaging, and viewer response and turns them into a measurable next-video test.
Questions readers usually ask next
What YouTube metric should creators prioritize?
There is no single best metric. Use the metric tied to the current question: reach, click-through, viewing quality, returning audience, or subscriber outcome.
Should Shorts and long-form videos be compared together?
Usually analyze them separately first because distribution, duration, and retention behavior differ.