The most important lesson I had when building all these products is that “If you don’t have product-market fit, nothing else matters”.

Finding product-market fit requires a different mindset and skillset than 99% of jobs: rather than executing and optimizing, it’s more about discovery. It’s more like a detective or lab scientist than like a manager or a chef.

There’s no recipe to follow and you can’t just play the game well and win. It’s about navigating a complex unknown and uncertain space that’s always changing, and you often don’t know when and what you’ll discover next.

Detectives can’t know when they’ll solve a crime, and lab scientists can’t know when they’ll discover the next cure, finding product-market fit, unfortunately, follows this pattern.

Under my product studio, I built and launched 6 consumer Apps:

• Cocina Feliz: What started as a way to learn Flutter became my first end-to-end effort to make money selling Apps. Here’s a promotional video. I learnt that the closest alternative (going on YouTube or Google) was pretty good and free. I found myself in the content marketing space and I wasn’t interested in growing a content marketing startup in Mexico, so I moved on.
• Purpose → Goalbook: What started as a two week project to capture goals and track progress on the phone (think OKRs), became an exploration into the productivity space. We tried a chat agent using conversational DialogFlow (before LLMs) to be like a coach, but we found this was too complicated and sometimes a UI form is better than a conversation, so we changed into Goalbook. We were having a hard time selling Goalbook, because we found that people would enter at most 3 goals at a time, confused tasks with goals, and the value of goal setting and following comes over the long-term, so most people aren’t able to capture the value immediately or even asses the impact, so they weren’t ready to pay for that, besides, a simple note would often do the job well enough.
• nemo: This way a quick 2 week “warm-up” project to build an App that would allow me to focus better on one thing when working on the computer. The concept was simple, the phone displays a task full-screen and you place the phone on your desk where you can see it, connected and with the screen on, this way you come back to the active task easily. The problem was remembering to use the App, and taking the time to fill in the task and position it correctly. Maybe if you could control it from the computer that could work…
• LEEST: For Goalbook to handle tasks (basically adding to-do list functionality) we had to almost reinvent to product, so we came up with this idea of list-based organization where you have lists of different types (goals, urgents, tasks, etc..) and each list behaves different, for example, the urgent list sends you a reminder every morning, the goals list checks-in on you from time to time, etc. The most liked feature is automatic AI-stamped emojis as bullet points, and the UI/UX is very polished. We love LEEST, but we ended up with a to-do list App with a “better” UX, and no new innovation. We did try a lot of things, like daily AI agendas via assistant chat, but none of that was disruptive enough to justify the investment to enter a saturated market, so we moved on. At some point I brought in a COO with the help of a financial advisor, but the bottleneck was the product, there wasn’t much to manage until we solved product discovery, she helped me conduct interviews based on a framework that tries to capture jobs-to-be-done. We also tried branding it as a to-do list for women, as Juliane suggested, it created controversy and some liked the idea, but we were met with criticism around exclusion and biased assumptions on gender roles, so we moved on. 10 people having been using the App every week for 3 years! We just didn’t think it was worth growing and trying to monetize on this.
• TomeGenius.com: I started playing with OpenAI before it was chat and I thought maybe it would be now possible (with the coming of DALL-E with good enough image generation) to use the format of social media feeds but to encourage learning. I’ve been testing it and I found that most people want something different, but also don’t want to pay for custom content (expect content to be free on the internet), and most people don’t like to learn for fun and without merit (other entertainment alternatives are more fun).
• Shape.it: A beta tester for TomeGenius saw one of my custom AI tools for TomeGenius.com and was very interested in using it, he offered to pay, so I vibe coded a prototype, validate it with him, and found that he really wanted that, but it was very hard to use, so I went and built a solid product from the ground up. I will be launching this product soon.
• Cuts.pro: After testing the possibility of making physical diagram cards for Montessori schools (because Nano Banana 2 was already very good with text), school directors actually said they didn’t need that, but they did need something like a “Canva for schools” so I understood the problem deeply and over 3 months, I took on the challenge of building “a design tool for classrooms”, I’m now working on my go-to-market strategy, getting users, and growing. I vibe coded this App with an interesting stack: Unison Lang + Cloud for backend and Flutter Web for the frontend.