As I mentioned in the previous post, I started researching a new niche, one that I really love: food and gastronomy. Going to different restaurants every week with my fiancée is one of my favorite hobbies, and we’ve also done gastronomic trips together. Our last one was to Belém, in the north of Brazil, which has an incredibly unique food culture unlike anywhere else in the country. Everyone who knows me knows this about me and always asks for restaurant recommendations, and I can never remember all the places I’ve been, and I don’t always take photos to help me remember what I ordered.

So I decided to validate this niche: interview people, understand the competition, and see if the pain was real.

This time the result was very different from the previous niche. I talked to people with very different profiles, and what became clear is that the pain isn’t universal. It shows up strongly in people who go out frequently and act as a food reference in their social circle. Someone who goes to a restaurant once a month doesn’t feel this because they end up remembering the few places they’ve been. But someone who goes every week, does food tourism, and constantly gets asked “where should I eat?”, that person feels it a lot.

The pattern that came up in almost every conversation was the same: everyone had already tried to solve it somehow. Spreadsheets, Notion, iPhone notes, Google Maps favorites. And everyone stopped using them. The most common reason was friction in the moment: you’re at the restaurant, enjoying the experience, and when it comes time to log it, it’s too much of a hassle. So you leave it for later, later never comes, and the memory disappears along with the bill.

This gave me a very clear insight about what the product needs to solve before anything else: logging has to be fast. If it takes more than 30 seconds, it becomes just another tool people will abandon.

With the pain validated, it was time to make decisions.

The name

After a lot of thinking, I chose Savor. It means to enjoy something slowly and intentionally, savoring the moment. It captures exactly what I want the app to transmit, and it works naturally in Portuguese and Spanish too, which are the other two languages the app will support.

Platform

I decided to go native iOS with Swift and SwiftUI. I had never built anything for iOS before I’ve always used React Native, so to learn I built a small personal finance app this week before starting Savor for real. It helped me understand the iOS architecture, how Xcode works, and I even built a widget, so Savor will have a widget too.

The vibe I want for the app is dark and refined. People who go out to dinner every week tend to have a more sophisticated lifestyle, and the app needs to reflect that. I also love the Liquid Glass aesthetic in iOS, which was one of the reasons I went native.

How I planned the product

Before writing any real code, I used a skill called grill-me that asks rigorous questions about the product, one at a time, until all the important decisions are locked in. It helped me think through edge cases I wouldn’t have considered on my own, things like: what happens when I visit a place that was on my wishlist? What happens when I go back to the same place a second time?

Three rounds of decisions came out of it: what to build and with what, how to build it, and the name. In the end I documented everything in a CONTEXT.md file that any AI agent (or future me in three months) can read and immediately understand where the project stands and why each decision was made.

The MVP in three phases

• Phase 1 - Habit: quick log and wishlist. The quick log only requires a name and a rating, everything else is optional. This is the 30-second rule in practice.
• Phase 2 - Memory: map with pins and visit history.
• Phase 3 - Viral: Instagram sharing and public web links.

What’s next

Next week I’m traveling to NYC and Toronto, two of the biggest gastronomic capitals in the world, and that’s the deadline I set for myself to have the MVP working. The focus will be validating phases 1 and 2: the quick log, the wishlist, and the map. The timeline is tight so the sharing features probably won’t make it in time, but that’s not what needs to be validated right now. The core of the product is logging fast and remembering later, and that’s what I need to test in practice, with real restaurants.

Before I leave I’ll share a demo showing how the app looks. And after the trip I’ll be back with a full overview of what the experience was like using Savor in the real world.