The one decision that determines whether you are still doing this in six months.
Why the niche decides everything
Not editing. Not the algorithm. The lane.
Almost nobody quits because their edits were bad. They quit because they picked a subject they thought would perform, ran out of things to say about it in five weeks, and could not face a sixth.
The test is not 'does this get views'. The test is: could you make forty videos about it without being paid, and would you still be interested at the fortieth? That is why the programme asks you to say your lane out loud, in public, before it gives you a link.
A lane you can survive beats a lane that performs, every time, because the one that performs is also the one everyone else is in.
How to read a niche before you commit
Saturation, RPM and whether it tolerates an affiliate offer.
Three questions, in this order. Who is already doing it well, and how long did it take them? What does the audience actually buy — because an audience that buys nothing monetises at ad rates and no better? And can you recommend something honestly, or would every affiliate link feel like a betrayal?
The third question is the one people skip and the one that ends channels. If you cannot imagine an honest recommendation inside your lane, you have chosen a lane that pays in views only.
We are not publishing an RPM table until we have measured one ourselves across enough channels to mean anything. Numbers copied from someone else's blog post are how this industry lies to you.
GPT-Video Partners pays a share of subscription revenue from customers you refer, and — once you have sold to customers of your own — a capped share of the revenue from customers referred by partners below you. It is not an income opportunity, a job, or an investment. Most participants in affiliate programs earn little or nothing. Nothing on this page is a promise, projection or guarantee of earnings. Any figures shown are actual results from named participants and are not typical.