Hey IH 👋
This is the story of a project I killed too early, why I brought it back, and what I'm building now.
A while back I launched AnimeMyPic — an AI tool that transforms your photos into anime-style art. Upload a photo, pick a style (Ghibli, Naruto, One Piece, Dragon Ball, and more), get a high-quality anime version of yourself.
The first version actually worked. Not perfectly, but it worked — and more importantly, people were paying for it. Not life-changing numbers, but real people, with real credit cards, buying credits to transform their photos. I had zero ads. Zero posts. No marketing whatsoever. Just organic traffic finding it somehow and converting.
That felt good.
Then the Supabase free tier hit its limit. I looked at the numbers, looked at the renewal cost, and made the classic indie hacker mistake: I decided it wasn't worth it. Shut it down. Moved on.
A few months passed. I wasn't thinking about it. Then I opened Vercel analytics — just out of curiosity — and saw something that stopped me.
People were still coming.
Organic visitors. Not a flood, but a steady trickle. People who had found the site through search, clicked around, tried to sign up, hit the auth wall, checked the pricing page. They wanted to use it and couldn't because I'd pulled the plug.
That was enough for me to decide to relaunch — properly this time.
Transformations now complete in 25–40 seconds. Images are compressed and resized client-side before being sent directly to the API. The result only gets saved after the transformation succeeds — nothing is stored unless it worked.
New feature: turn your anime transformation into a collectible trading card styled to the universe you chose. One Piece gives you a wanted poster. Naruto gives you a ninja dossier scroll. Yu-Gi-Oh gives you a proper duel card layout. Rarity (Common / Rare / Epic / Legendary) is assigned deterministically per user.
Full gallery of your transformations with sharing, download, 4K upscale, and card generation — all in one place.
Monthly credit packs via Stripe for people who transform regularly.
Proper forgot password flow, smoother onboarding.
The organic traffic never really stopped even while the site was down. That tells me there's something here worth building on.
No ads. No paid posts. Just search traffic and word of mouth — same as before, except now the product is actually ready for them.
If you want to see yourself as a Ghibli character, a Naruto ninja, or a One Piece pirate — give it a try.
And if you've ever shut something down too early — maybe check your analytics first.
Would appreciate any feedback
I took a look at the site, and I think the biggest opportunity is not the image quality, it’s purchase confidence.
A few things that would likely lift conversion fastest:
The hero tells me what the tool does, but not why I should trust the result enough to upload a photo and pay. I’d put 3 strong before/after examples above the fold immediately, ideally with recognizable style labels and one "started from this photo -> got this output" flow.
"Join thousands of happy users" feels risky when the page still says "Loading reviews...". Right now that undercuts trust. I’d either ship real testimonials/screenshots or remove that section until it’s real.
The page is feature-rich, but the buying moment is emotional. I’d make the first CTA more outcome-led, something like "Turn your photo into anime art in 30 seconds" rather than generic transform language.
If organic traffic is the wedge, I’d add a visible pricing/credits hint earlier so visitors know this isn’t a bait-and-switch after they upload.
The trading cards / backgrounds / 4K extras are nice, but they may be distracting the first buyer. I’d sell the core transformation first, then upsell the extras after the first successful output.
If useful, I also run very fast $1 homepage teardowns for indie founders, focused on what is most likely to increase clicks and paid conversions: https://roastmysite.io/manual-payment-link.html?src=external_manual_ih_animemypic_confidence_apr23_usd_presell_hv