Knowledge Bank
3 min read
Video stats, and the thumbnail history nobody checks
Lifetime numbers on any public video, plus the thumbnails it has worn, which is where the real story usually is.
Not sure if Plan is your weak phase? The free 3-minute workflow audit will tell you.
Scout the competition is step 3 of the 47, and the usual version of it is watching five videos and forming an opinion. Video stats gives you the numbers underneath any public video instead: lifetime figures, an estimated revenue band, a view timeline, and the thing almost nothing else shows, a history of the thumbnails that video has worn.
The thumbnail history is the interesting part
Thumbnail changes are pulled from the Internet Archive, which means you can see the images a video has carried over its life rather than only the one it is wearing today.
This changes what you learn from a competitor. You stop asking why that thumbnail worked and start asking what the first one got wrong, which is a far more useful question, because the first one is the version of the problem you are currently in.
Read the timeline against the changes
| What you see | What it probably means |
|---|---|
| A steady climb from day one | The packaging worked immediately. Study the first thumbnail |
| A flat start then a jump, with a thumbnail change at the same point | The idea was fine and the packaging was not |
| A jump with no thumbnail change | Something external happened. Suggested traffic, a share, a trend |
| A spike that decays fast | The click was earned and the video did not hold people |
The revenue estimate needs a niche
You pick the niche yourself, because the same view count earns wildly different amounts depending on who the advertisers are trying to reach. A number without that context is a number pretending to know something it does not.
Treat the output as a band rather than a precise figure. It is useful for working out whether a format can support the effort it takes, and it is not useful for guessing what a specific creator earns.
Your own videos get more
On videos from your connected channel, the view timeline comes from Studio daily views rather than being inferred, so the shape is exact rather than estimated.
That makes the same tool a debrief instrument as well as a research one. Run it on your last upload at 48 hours and you get the curve next to the packaging that produced it.
What it will not tell you
Retention. None of this reads how long people stayed, and a video that pulled a million views and lost everybody at ninety seconds looks identical here to one that held them.
Views measure the click. Watch time measures the video. Use this for the first question and your own analytics for the second.
Keep reading
The wider approach is covered in competitor research, and comparing two videos properly is in the video comparison snapshot. When you are ready to judge your own, reading the first day's numbers is the calmer version of that job.