I spent 2,000€ on my first app with 0 conversions. Rebuilt with AI at the core — first paying user in hours.

Strong pivot. Building MetricSync taught me something similar: the magic is not the model, it is removing the second decision.

For FoodCraft I would watch the moment after the recipe suggestion. If users start editing calories/macros, servings, or ingredients before saving, the AI answer is probably too brittle.

I would make the first run very forgiving: photo or text dump of ingredients, rough goal like high protein, then one-tap corrections.

Also, your ad lesson is useful. The winning angle may not be "AI recipe app". It may be "I have random food at home and need dinner now."

Love the pivot 🔥

One interesting marketing angle could be “Maa ke haath ka khana❤️”

In many cultures (especially India), people don’t want fancy recipes ,they want traditional home food.

What if users could select regional/traditional dishes and the AI adapts them to their protein targets and calorie goals without changing the taste profile too much?

That emotional + health combo could be powerful. Would love to see something like that inside FoodCraft 👌

If you want, i can help you with this😊

Congrats on the turnaround, that must have felt really validating after the first version.

We just launched our own consumer app this week after building it for about a year on the side, also with AI playing a big role in making it possible for a small non-technical team to ship something complex.

One thing we’re already noticing is that the biggest difference isn’t just the AI feature itself, but how it changes the way users interact with the product. In our case, speaking instead of typing seems to create a very different level of engagement.

Out of curiosity, did your first premium user come from someone who used the personalized features immediately, or after exploring the app for a while?

That 2,000€ lesson hits hard but you learned fast. Most founders keep throwing money at broken funnels instead of fixing the core problem.

The semantic search angle is brilliant. I'm working on mobile apps and noticed something: people keep their phones in the kitchen while cooking. Have you considered how FoodCraft works on mobile? Features like voice input for ingredients ("I have some chicken and whatever's in the fridge") could be huge since people's hands are often messy while cooking.

Also curious about the macro adaptation piece. Are you finding users prefer exact macro targets ("I need 30g protein") or more flexible goals ("high protein, low carb")? That could affect how you position the AI - precise tool vs flexible assistant.

Either way, congrats on finding real demand. That immediate conversion after rebuilding tells you everything.

The 2,000 euro lesson is one most founders learn the hard way. You can't outspend bad product-market fit. The fact that 3 euros converted better than 2k tells you everything about what changed.

Using pgvector for semantic recipe search is a nice touch. "Chicken, some rice, and whatever that green thing is in the back of my fridge" is genuinely how people think about cooking, not browsing categories. That's a subtle but important UX difference from traditional recipe apps.

I'm curious about the GPT-5 vision analysis part though. Are people actually snapping pics of their fridge and getting recipe suggestions? If so, that's a really sticky feature because it removes the friction of even typing ingredients. If not, might be worth pushing that harder in onboarding since it's the kind of thing that creates a wow moment.

The meal plan + auto shopping list combo is where I'd bet the retention lives. Individual recipe searches are one-off, but weekly meal planning is a habit. Good luck with the growth!

Great pivot story. I'm also building in the AI space and the difference AI makes in time-to-value is insane. What was your biggest lesson from that first version?

The rebuild with AI is smart. Sometimes you have to burn through the first version to really understand what customers want. What was the biggest thing you changed between v1 and the AI version?

Love the simplicity and clarity of FoodCraft’s value proposition, “Eat what you want, we crunch the numbers.” That’s a really approachable way to explain complex nutrition without sounding technical.

The way you use AI to tailor recipes to individual calories, allergies, and diet goals feels very practical and the 3,200+ adaptable dishes look like a great resource for users looking to cook and stay healthy.

What stood out to me most was the combination of personalized meal plans with automatic shopping lists that’s a real pain point for people juggling fitness and nutrition.

One suggestion: have you considered adding a progress tracking dashboard? It could help users see how their meal choices affect long-term goals (like weight loss or muscle gain).

Great rebuild story - the jump from 2k euros with 0 conversions to 3 euros with 1 paying user really shows how much more important product-market fit is than ad spend. The AI ingredient matching is clever since most people do open the fridge and wonder what to make. Curious what your retention looks like after the first week - do people keep coming back for meal plans or is it more one-off recipe searches?

Great story ! Congrats on the launch ! Just wondering : 3€ is the total budget you spent or its a price per one signup?

The 2000€ to 3€ conversion story is a great lesson in finding the right problem first. Most of us (myself included) start with a solution looking for a problem. You flipped it - started from a real pain point and built around it. The pgvector semantic search for recipes sounds like a genuine technical moat. Curious: are users mostly discovering recipes through ingredient input or the meal planning flow?

Great concept! Simple, clean design, and the focus on healthy, homemade meals makes Foodcraft appealing and trustworthy. Looking forward to seeing its growth!

The 2000 euro to 3 euro pivot is the real story here. Most founders keep throwing money at distribution when the product itself is not solving a specific enough problem. The fact that you went from generic recipe site to AI-powered personalization shows you actually listened to users. One question - are you seeing people come back daily for meal planning, or is it more of a weekly use case? That retention signal will tell you everything about product-market fit.

The 2000 euro to 3 euro pivot is the real story here. Most founders keep throwing money at distribution when the product itself is not solving a specific enough problem. The fact that you went from generic recipe site to AI-powered personalization shows you actually listened to users. One question - are you seeing people come back daily for meal planning, or is it more of a weekly use case? That retention signal will tell you everything about product-market fit.

Congrats on the launch, looks solid. How are you currently thinking about acquiring early users and gathering feedback?

2,000€ → 0 vs 3€ → 1 conversion. That's the brutal honesty every founder needs to see.

You didn't just add AI as a feature — you repositioned around a real pain point. "I have food but no idea what to make" is a daily frustration. Recipe sites just give you more options; you gave a solution.

The semantic search with pgvector is smart — "I have chicken, rice, and something green" is how people actually think, not "show me chicken recipes."

One question: how are you handling the "I'm feeling lazy" use case? Like, I have ingredients but I also don't want to cook for 45 mins. Is there a "effort level" filter in there? That feels like the next layer of personalization that could drive retention.

The 2,000€ → 0 conversions vs 3€ → 1 conversion contrast is the clearest signal you could ask for. You weren't solving the wrong problem badly—you were solving a problem that didn't exist strongly enough.

"I have food at home but no idea what to make with it" is one of those universal friction points that people feel multiple times per week. The fact that your first test spend converted immediately suggests you've hit actual demand.

The interesting part is what happens next. That first paying user chose to pay within hours—what specific moment made them convert? Was it seeing their exact macros match a recipe they already wanted to make? The meal plan solving their weekly decision fatigue?

Understanding that precise conversion moment will tell you whether you're selling convenience (meal planning), control (macro adaptation), or solution to waste (using what's already there). Each of those attracts a different user, tolerates different pricing, and has different retention patterns.

Would be curious what that first conversion actually valued most.

Huge difference between “another recipe site” and solving the actual friction.

The fact that 3€ produced a premium conversion while 2,000€ didn’t tells you the positioning shift is real.

One thing I’d be curious about: where is the strongest demand cluster forming right now (calorie tracking crowd, busy families, or macro-focused fitness users)? That segment focus could massively amplify your early traction.

Would love to see which audience is responding most.

thats a crazy turnaround from 2000 spend to first conversion.

which feature is pulling best right now, recipe adaptation or weekly plans? are you seeing repeat usage after day 1?