Test and Iterate
Treat every video as an experiment, read the analytics, and do more of what works.
- Difficulty
- Easy
- Time to result
- ~ongoing to results
- Steps
- 3
- Confidence
- 85%
Test and iterate is the core operating philosophy behind the Try Guys — the name itself reflects that they do everything as a test and keep what works. Fulmer frames every video as an experiment, whether it's testing a visual style, a content topic, or a new cast member. After release they study the analytics to gain a concrete learning, then produce more of what performed and less of what didn't. In practice the main levers they A/B test are titles and thumbnails, running two variants against each other. This turns content creation from guesswork into a compounding feedback loop.
Origin
The Try Guys name literally came from their working method: they would try something, see if people liked it, and if so do more of it — there was no top-down branding decision.
Core principles
- 01Every video is a test — of style, content, or casting.
- 02Read the analytics after each release to extract a learning.
- 03Do more of what's working and less of what isn't.
- 04A/B test the levers you actually control, primarily titles and thumbnails.
How to run it
- 1
Frame every video as a test
Before publishing, define what the video is testing — a style, a content angle, or a casting choice — so there is a clear hypothesis to learn from.
- 2
A/B test the levers you control
Run two thumbnails and two titles against each other to see which drives more clicks, since these are the elements you can genuinely vary on YouTube.
Watch out Keep title and thumbnail honest to the content — a mismatched hook that says something the video doesn't deliver just confuses viewers.
- 3
Read the analytics and adjust
After release, study performance to extract a learning, then make more of what worked and less of what didn't.
In the wild
The Try Guys create two different thumbnails and two different titles for a video and play them off against each other to see which performs.
→ They identify the higher-performing variant while keeping the messaging consistent with the actual content.
Common mistakes
Testing with a misleading title
An A/B title that promises something the video doesn't contain confuses viewers who click through, undermining the test.
Is it for you?
Best for
Channels with enough volume to run experiments and read performance analytics.
Not ideal for
One-off projects or creators with no access to platform analytics.
From the transcript
“we think of every video as a test and then we look at the analytics we try and gain learnings from it whether it's a…”
“do more of the stuff that's working and do less of the stuff that's not”
From the episode
Ned Fulmer: How to Go Viral on YouTube
Ned Fulmer