We actually shipped an update to the ranking algorithm today that addresses exactly those sudden drops of thousands of positions. Rankings should now decline more gradually, giving creators more time to react instead of seeing their templates disappear overnight. So yes, we’re listening, and we’re actively working on improving things.
On AI, I don’t think it’s as black and white as “AI-generated template = ban”. We believe AI, when used properly, can be genuinely useful, and there’s a much broader conversation to be had about where to draw that line.
On quality checks, I agree that giving creators better tools to catch issues before publishing would be valuable. And as you pointed out yourself, AI can already help with things like detecting broken links and other common mistakes.
That said, we also believe creators should take ownership of the quality of what they publish and the experience they provide to buyers. We’ve spent years manually reviewing templates and helping creators identify issues, but that approach becomes increasingly difficult to scale as the Marketplace grows. The goal is to find better ways to support creators without making them dependent on a manual review process.
And as always, we’re always open to constructive feedback. Seriously. A lot of the improvements we make come directly from creators sharing their experiences. But there’s a big difference between explaining what isn’t working, why it matters, and suggesting improvements, versus repeatedly saying everything is broken and demanding we fix it.
We want the Marketplace to succeed just as much as you do, and we’re putting a lot of work into making that happen.
Of course we know how our own ranking algorithm works. That’s the least you should expect from us.
Building a ranking system that feels fresh every day, gives everyone a fair chance to reach the top, and consistently surfaces quality work isn’t something you can build in a week or just prompt Claude to figure out.
There’s a lot of experimentation and iteration happening behind the scenes, and we’re fully aware that we’re still far from getting everything right.
But as you can probably understand, we deliberately don’t share the exact mechanics of our ranking algorithm. Making those details public would make it significantly easier to game the system, which would ultimately hurt the very creators we’re trying to support.
So yes, there’s a degree of obscurity around how ranking works, but that’s a deliberate decision, not because we don’t understand our own system.
We’re listening to the feedback, we’re constantly making improvements, and we absolutely want creators to feel that their work has a fair chance of being discovered. That’s what we’re working toward.
@shivangshah_ Nothing stopping you! But aren’t you essentially using those 3-4 websites as templates then? 😄
That’s kind of my point. Instead of having an agent recreate something from visual references, why not start with an actual working website and personalize it from there?
Hey! These aren’t randomly picked. We use a service that analyzes the entire video and automatically selects what it considers the best thumbnail.
We know the results aren’t always ideal, so we’re exploring ways to let creators either choose a specific frame or upload their own thumbnail.