How I went from idea to product to killing it all in the span of a week

It started with scrolling Instagram at the gym. I came across a video of a photographer in New York waving a flash around and getting some really cool portraits on unsuspecting subjects. “Damn” I thought to myself “I need to start doing that” - and boom, just like that, inspiration.

It happens relatively often when the algorithm gets it right, and it's one of the major reasons I can't seem to get off of social media. You come across something tied to your interests that makes you stop and look at things slightly differently, pushes you to pursue something, or for whatever reason inspires you. A beautifully decorated home in a magazine, a launch post for a product on Twitter, a music tutorial on YouTube.

And so I did what I always do with this kind of stuff. I used Instagram's own bookmarking function and scrolled away, only to stop and realize I'm never going to look at that video again. All my saved posts scattered around different platforms, their sorting is messy and by the time they would come in handy I’ve usually forgotten about them. So that's when it hit me. What if I made a single place where I could chuck anything that inspired me, anything that I want to read later, any random thought that I had and these could magically be sorted for me and surfaced when most appropriate in a digestible way? I think there’s something there.

The first thing I did was check if anything like this already exists. I don't want to reinvent the wheel, and unless I can make something better than what's out there, it doesn't make sense for me to work on anything new. Pinterest was the obvious first thing to look at, but it is primarily intended as a visual medium, and it doesn't feel like the right place to add links from around the web, or my own random thoughts. Also, there was no feature to create any sort of briefing for me to digest the content. After this I played with a number of tools to save things from around the web, but they all had their gaps: Pocket is dead, MyMind doesn’t have any briefing function, and Sublime has too much emphasis on the social component.

Okay, perfect. I had an idea which had versions of it out there that are successful but that are missing touches I can give it. Let’s get started.

So over the course of a couple of days, my trusty companion Claude Code and I worked on this. The concept is simple: you add a link, an image, a thought, and Thimble (placeholder name) sorts it to the right category for you and creates briefings that group and summarize them, as well as making suggestions if any of the things you added are actionable.

It was ready. I hosted it on Vercel and decided that the litmus test was going to be whether I or any of the friends I shared it with would find it useful in the next few days. And at first it was exciting: I found myself adding things to it that I wanted to go back to, things I wanted to read later, random thoughts I had. And then I’d go back to my saved items and revisit them. However, the main sell for this app, the differentiator from what is already out there, the AI briefing, wasn't panning out. No matter how many iterations I went through with different prompts, different grouping techniques, how much data was scraped or ways to pass it the data, the briefings that were being generated were just not useful. My friends also never touched it, which in itself is excellent feedback.

It became clear within a few days of using it that no amount of AI briefing was going to pull me back to things I had saved if I wasn’t already interested in going back to them myself. And if that was the case, the briefing would just end up adding fodder to the content I saved which I just wanted to go back to. I played around with the idea of summarizing all the saved content and offering it as a wall of text, but this fell victim to the same problem as before. I thought about doubling down on suggesting new content based on the saves, but at that point I'd be competing with the sources of this content (Reddit, Twitter, Instagram, etc.) which do a much better job in surfacing and suggesting content. I also thought of creating a personalized home page/newsletter where all your saves would be served to you as a magazine, but like I said before, no amount of reshuffling content is gonna make you want to look at it, and if you already want to look at it, you might as well serve it in the cleanest possible way.

So, for now, I killed it. There's plenty of ways out there to save stuff and come back to it later, and it turned out that that was the most useful thing Thimble did. I could have probably validated the whole thing within a day by just grabbing a ton of thoughts and links and chucking them into a Claude session to see what it did with them, but then I wouldn't have been able to design a beautiful UI/UX, which is one of my favorite things to do. Onto the next one.