Case study
THE CHANNEL
A channel with a high floor and no obvious formula
AI LABS is run by autometa.dev, a software company. The channel has 174K subscribers and 287 videos, and the median of its last 30 uploads is 52K views, a strong floor.
The top six are 209K to 489K. At first glance they look unrelated: Cursor, Claude, Figma, a coding method, a GitHub tool, an n8n workflow.
Look closer and the list sorts itself.
The top six
3 Ways to Build ACTUALLY Beautiful Websites Using Cursor AI: 489K
Insane Claude Design Skills You Need To Actually Build Beautiful Sites: 372K
How to Use Figma AI To Make Sites 10X More Beautiful: 257K
The BMAD Method: The Ultimate AI Coding System: 251K
GitHub Top Trending Tool Just Fixed The AI Agent’s Biggest Problem: 230K
This n8n mcp is INSANE… Let AI Create your Entire Automation: 209K
WHAT THE DATA SHOWS
The pattern: one pain, a named tool, a promise of “actually good”
1. “Beautiful” is the pain
The top three videos (Cursor, Claude, Figma) are the same promise: AI makes sites that look generic, here is how to make them look designed. That is a pain a viewer feels every time they ship something, so it keeps getting searched.
2. The tool name does half the work
Each title names a tool people already use or want to learn. The viewer is not asked to care about a new idea. They are asked to get more from a tool they have open in another tab.
3. Recency is part of the format
Last week’s uploads (“Google’s New Way To Use Claude Code Is Insane”, 113K in a day, “Google Just Fixed AI Design”, 67K) are riding the same news cycle. A channel that is first with a clear take on a new release gets the early search and recommendation traffic.
4. Numbered, specific promises
“3 Ways”, “5 Use Cases”, “6 Insane Trending Repos”. The viewer knows the shape of the video before they click.
WHAT WE DO WITH IT
What we do with it
Make “design quality” a series. One tool per episode, same title skeleton, same structure, so the audience knows what the next one will deliver.
Build a release calendar. A list of tools and models likely to ship updates, with a 24-hour take ready to record.
Add the product thread. Software built by autometa.dev appears inside the lesson as the thing being built, not as a pitch.
Track by pain, not by tool. A tool can fade; “make AI sites look good” has not.
Questions a SaaS founder would ask
How do we know a trending topic will still work in three months? Check whether it solves a problem the viewer has every week. “Beautiful sites” does; a single release does not.
Should we chase every release? No. Pick the ones that connect to your product’s job, and take a clear position.
Why does one tool name outperform another? Size of the tool’s user base and how new the release is. Compare against the channel’s median, not another channel’s.
Source: public YouTube data for @AILABS-393, captured 27 October 2026. Median is the middle view count of the last 30 uploads.
AI LABS
Industry
AI development tools and education
CHANNEL
174K subscribers · 287 videos
WEBSITE
autometa.dev
Want results like these? Book a free discovery call and map your YouTube opportunity.
More channel breakdowns
SEO software
Education
324K views from one course. Veloifs helps Surfer Academy repeat it.
324K
Views on the top video
64×
Top video vs median
35.3K
Subscribers
Read success story
↗
Fintech
Banking
678K views without mentioning a bank account. Veloifs helps Rho repeat it.
678K
Views on the top video
48×
Top video vs median
34.1K
Subscribers
Read success story
↗
AI tools
Education
489K views on one pain. Veloifs helps AI LABS turn it into a series.
489K
Views on the top video
9.4×
Top video vs median
174K
Subscribers
Read success story
↗