Overall, it increased downloads and boosted impressions, but there are still many other factors I need to consider 🤡.

You start to notice common patterns in how they attract users’ attention—the visuals of the food, the order of the content, the color scheme, and the messaging. Ultimately, I ended up with the following design.

While talking with my friend, I realized how much a good first impression influences a user’s decision. I learned a lot by researching how competitors market their apps.

I went through five iterations, each with a different focus, ranging from copywriting and marketing to design.

With a web development background, I didn’t take App Store screenshots very seriously, but experimenting with different designs taught me some valuable lessons.

Many apps have a chat interface, but it can be slower then correcting the input value itself. So, a combination of chat interface and swipe gestures was chosen. Users can swipe to delete without an additional call to the AI API and can prefill the chat with a quick quick click.

I am also experimenting with the UX to find a ideal balance between the convenience of AI and the correction mechanism. Since it can give wrong answers I wanted to have a user friendly way to quickly correct it and also have a diverse option to correct anything.

You also need internet connection to use it, which is annoying when you want to log frequent meals from the past.

I did not like other AI calorie tracking applications like Cal AI, Eato or meals chat, because they lacked core features you would expect from typical food trackers (Saved meals, change date and time, portion adjustment based on serving / weight).

I was tired of not gaining weight, so I tried few calorie tracking applications. The closest one I liked was Macrofactor, but they did not have a feature to snap a photo (I tend to eat out outside often). So I decided to create my own application. #buildinpublic