Gapspot

That's rare for early-stage projects. The homepage is clean, intentional, and doesn't scream "I let ChatGPT design this." Someone clearly thought about guiding you toward the main action: trying the demo. Navigation makes sense, the value prop is clear, and the design feels like a real product.

But I walked away from the homepage thinking there's no free trial. It's not in the hero section, it's not in the pricing. For a tool that needs you to experience it before you buy, that's a big miss. Turns out free access does exist, but I only found that out inside the demo. That disconnect needs fixing yesterday.

The demo lets you run an action on a pre-defined subreddit. Sounds fine, except I ran it multiple times and got the same response every time. That screams pre-generated. If I'm evaluating whether this tool is smart enough to pay for, showing me a canned answer is the worst possible move.

A few other demo issues:

The demo didn't convince me to subscribe. It didn't show me how the real product works, and the limitations made it feel like a toy.

Registration and login work smoothly. Once you're in, you get free tokens, nice touch. I tried both the preselected categories and a random subreddit.

The preselected categories felt pre-generated again. That creates a problem: if I pick an idea from there, I'm probably picking the same idea as everyone else who tried the platform. That's the opposite of finding a gap in the market.

But the random subreddit search? That actually worked really well. It took a few minutes, but there's a banner explaining the wait time, good UX decision. The results felt genuinely useful and tailored. This is the strongest part of the whole product.

The free tokens run out fast. I exhausted my tries quickly, and by that point I wasn't convinced enough to pay for more. The random subreddit feature gave me hope, but one or two good results isn't enough to pull out my card.

I'd give users more tokens on the free tier. Enough to generate a handful of real ideas. Let people see it work three or four times before asking them to pay. If the random subreddit results are consistently that good, people will convert on their own.

The idea is worth pursuing and the execution for an early project is impressive. The good stuff is just buried too deep right now.