YouTube has announced new Research and Insights destinations in Studio, including a way to discover outlier videos. Here’s what the announcement confirms and a practical way to use that kind of research when planning your next video.
This article analyses YouTube’s announcement. We haven’t tested an account rollout of these destinations.
What YouTube has confirmed
In its 23 September 2026 announcement, YouTube describes Insights and Research as destinations for understanding past performance and finding outlier videos that can inspire future content.
The announcement doesn’t provide a complete outlier-scoring formula or a detailed eligibility list for those destinations. Check your Studio account and current official guidance to see what’s available to you.
The important question is what to do with a signal
Finding a successful video is a starting point. The next job is to work out whether it reveals a viewer need you could serve.
Imagine you find a video about restoring an old laptop. It has done unusually well for its channel. Perhaps viewers need a cheaper alternative to a new computer. Perhaps the device is popular, the test is unusual, or the title makes a particularly clear promise.
Views alone won’t tell you which explanation is right. Write down a question you can investigate: “Do these viewers want to know whether an older laptop can still handle everyday work?”
That question gives your research a direction.
First, check the comparison
An outlier video stands out against a baseline. Before treating one as a promising idea, check that comparison:
- Format: are you comparing long-form videos with other long-form videos?
- Age: have the uploads had similar time to collect views?
- Channel: has its topic or audience changed recently?
- Typical results: does this video stand out against several comparable uploads, or just one weak result?
- Circumstances: was there an event, collaboration or other unusual reason for attention?
When you only have current public view counts, note that limitation. They don’t reconstruct a video’s first-week performance.
Then, identify the promise to the viewer
Describe what someone expects to get from the video in one sentence.
For the laptop example, that could be: “Find out whether the machine you already own can do the work you need.” A repair guide, benchmark test and buying comparison could each address that need differently.
Write down the promise you see in the title and thumbnail. Then watch the video and check how it delivers. This is your interpretation of the packaging, not proof of why viewers clicked.
Look for a pattern you can explain
Find a few relevant examples across related channels and topics. Do they support your explanation? Are there similar videos that performed poorly? If only one exceptional creator makes the format work, that matters too.
Keep the research manageable. Stop when you can explain the viewer need, the evidence behind it, the main uncertainty and what you would add. If something is missing, name the question your next round of research needs to answer.
Use this short research worksheet
Copy these prompts into your notes. The example is hypothetical; use your own observations when researching a real video.
| Prompt | Example note |
|---|---|
| What did I notice? | An old-laptop restoration video stands out among comparable uploads. |
| What am I comparing it with? | Similar long-form uploads on that channel, at comparable ages where possible. |
| What might the viewer need? | A way to keep working without buying a replacement computer. |
| What supports that idea? | Related videos address the same budget problem through practical tests. |
| What don’t I know? | I can’t see the channel’s private retention data or prove what caused the result. |
| What could I add? | Test a different machine against a specific set of everyday tasks. |
| What should I check next? | Whether I can get the device and run a useful test. |
A short note like this makes it easier to decide whether an idea deserves production time.
Turn the note into an original brief
Add four things to make a plan someone can work from:
- Audience: who is this for, and what situation are they in?
- Promise: what question will the finished video answer?
- Material: what test, footage, interview or explanation will you need?
- Uncertainty: what could make the idea unsuitable or difficult to deliver?
Then sketch a title and thumbnail. Check that the video you can actually make will deliver what they promise. If the title depends on a test result you haven’t confirmed, keep it provisional.
Questions worth keeping separate
Does finding an outlier mean I should remake it?
No. Study the need it serves and decide what original contribution you could make for your audience.
Can public views tell me why a video succeeded?
Not on their own. Views show an outcome. Explanations about timing, packaging and content need supporting evidence.
Should I buy another research tool immediately?
First identify a job you can’t do with the information you already have. You might need historical observations, faster comparisons or a better way to organise notes. Judge tools against that job.
What should I do today?
Choose one relevant video, check the comparison and complete the worksheet. Turn your notes into one original brief with a clear audience and promise.