The app I never launched
Back in mid 2014, I was fresh out of high school (I dropped out).
I would spend a lot of time on the internet learning about programming and technical topics, researching my interests, and I'd run into this issue repeatedly where I'd remember I'd read something interesting somewhere, I'd filed it away in my head as "intriguing, read later", but now it's been a week and I don't remember where I read it. It would really grate my nerves to be unable to find those articles and sites again, like I'd lost some part of myself and would feel empty until I got it back. I don't mean to be dramatic but it really did bother me a lot.
I used browser bookmarks, Pocket, Instapaper, Pinboard and dozens of other tools. All of them did one thing or another, but none of them did everything I wanted in such a service in one place. For instance, Pocket and Instapaper had this reader view to read articles without ads and other cruft, Pinboard had this "view in context" button so if you find a bookmark from search results, you could press a button and get taken to its page in the list, you can see what you bookmarked before and after it to get back that mental thread. Only a few had full text search, many did not tell me if something had already been bookmarked via their extension. I also had many instances of content being changed, removed, or the website itself disappearing, so I wanted archives which was a rare feature too.
I'm sure there were more, the point was, each tool I would try had one thing or the other, but none had everything I wanted in one place. What I wanted was a swiss army knife, and I realised, if I want this, I have to be the one to make it.
Back then, I had very little programming experience. I did believe that if I want this, probably others do too, and I could make a startup out of it. I was very active on hacker news, I used to watch silicon valley (HBO), I'd been inspired by pirates of silicon valley, I always knew I wanted to be an entrepreneur and this seemed like a great opportunity to do so. My idea behind dropping out had been that I will run and scale my own business. Getting a job had never been on the table.
I tried to get others to code it for me. I tried to hire an agency, but to be honest, I was broke (I was 16 or 17) and it was just wishful thinking that they'd agree to a revenue split or something. I tried to find other developers in our local tech meetups, but getting them excited about it wasn't easy. We all know how often developers are approached by people to build things for them (if you don't, it's very, very often to the point of being a tired trope). Long story short, I realised, if I'm gonna build this, I'm gonna have to be the one to code it.
This story is about how I never launched the app, but to be honest, in hindsight, it all worked out just fine. I also believe things don't just "work out", you have to make it so, and for that I am proud of myself for making it work out.
I used the project to motivate myself to become good at programming, learn about web development, browser extension development, async processing, relational databases, full stack development, deployment, scaling, performance, so many things. It formed the base of my knowledge for all I do today. It also compressed my learning journey, because I would learn whatever I needed to learn, when I needed to learn it. I have ADHD, I can't power through boring lecture driven education, where someone else has designed some curriculum, not based on what's most likely to be engaging, but whatever other metric has been used. I was able to do in-context learning, which also meant I immediately knew why something was important, and was able to immediately put learnings into practice. I've spoken about project based learning before How to learn programming. I think it's the best way to learn programming, maybe anything. For me, it certainly is. I learned for instance way more about playing the keyboard when I decided to learn to play the interstellar theme song, than when I paid hundreds of dollars for a course website. Buying courses has always been aspirational for me and never practical. I learn best hands-on, and this product provided the perfect playground for me to get my hands dirty.
You know, sitting here, when I started writing this post, it was going to be about how I failed to launch the product repeatedly, and I will get to that. But it's also turning out to be about appreciation for all it taught me along the way. I owe every dollar I've ever made to what started with this project. Could it have started differently? What couldn't. But this is how it started for me, and I'm proud of the journey I've had.
So anyway, I taught myself programming, and immediately would be applying my learnings to this app in parallel. I named the app Crestify, the name was born or a misunderstanding, I thought treasure chests were also called "crests" in some contexts? I started building the app, and I did math about how many hundreds of thousands of dollars I could make by selling it as a SaaS product every year. I had done research, Pinboard had their finances shared openly and they made similar figures. I was making a more featureful product, there's no reason I couldn't make the same. Pocket made millions every year, certainly I could make 10% of that.
I did not know what I did not know. That customers didn't grow on trees or drop out of the sky into your lap. Programming isn't easy, but getting customers is pretty hard too, especially when its something you've programming yourself. Customers, and especially the hacker news crowd I was involved deeply in those days, I knew would tear me apart. hacker news and product hunt were the only ways I really know "how" I'd get customers. Opening myself up to criticism and ridicule like that for something I poured my heart and soul into, something I stayed up until 5AM working on, something I'd worked on 14 hours a day for months on end. You can understand why that would feel scary and un-enticing.
I worked on the product, but I always told myself, "one more feature". I was like an addict. One more redesign, one more feature, one more rewrite and it'll be perfect to be released. Then people will just flock to it. When I launch it, I'm going to be overwhelmed by customers so I should already be ready to scale my backend 10x, 100x. I never actually replaced the original codebase, but I did rewrite it in Rails, Meteor, Django, I tried to rewrite it in Phoenix, probably others. They were all great learning exercises, they just missed the point that I wasn't trying to recreate the TodoMVC project.
By the end of it, Crestify had this long list of features:
- One-click bookmarking across many browsers
- Archiving of saved bookmarks to archive.org, archive.today
- Archiving of saved bookmarks on-platform, as you saw them (paywalled content, dynamic content, etc)
- Full text search over all your bookmarks
- Tagging
- Reader mode
- Saving of all open tabs and re-opening on another device / at another time
- Context view
- Saving of browsing history will full text search over paywalled, dynamic content
- Importing from Pinboard, Pocket, Instapaper, Readability, browser native bookmarks, and more
- Faceted search
- and more...
I built a feature factory. Any feature I wanted to add, I added it. I said no to nothing. The problem was that I would always get some idea for one more thing I need to implement before I press the button to take it live on product hunt etc. This was procrastination in disguise.
I finally abandoned hope of ever launching this product in late 2020 or early 2021, so 6 to 7 years after I started working on it.
Now before you misunderstand, this is not all I worked on for 6-7 years. I worked on it from 2014 to 2016. Then I cofounded other startups and worked there for the next 2.5 years cumulatively. One of these was quite successful and all the knowledge I built making Crestify came in really handy there. After that, in 2018, I left the startups and I moved to freelancing and being a digital nomad. From 2018 to 2020, that's what I did, until Covid killed that business in 2020. I then joined another startup for my first "real" job (ie, not my own startups). But throughout this time, I probably did spend 3-3.5 years tinkering with Crestify and never launching it.
And to be honest, I am glad I took Crestify into the backyard and shot it. At that point, it had fallen way behind. It still has unique aspects to it and it still does what it's meant to do quite well. In terms of UI and UX, it's fallen behind though. To the point I don't use it anymore myself. So, I realised, I would keep tinkering with this dead-end product, procrastinating my life away and killing all my dreams in the process, or I could set myself free from the obligation that just because I've worked on it for so long, I must launch it (but now the "just one more thing" would be UI / UX tweaks, then something else, so on and so forth...). If I "gave up" on this app, I could open myself up to newer, better ideas. And so I did. It wasn't easy to not feel like a huge failure, to not internalise all that negative messaging, to not consider the failure of a product idea to a failure of me the product builder. But eventually, I got over it.
I learned, again, so many things in-context. When they say, create a minimum viable product, launch early, launch often, if you're not embarrassed about your first version you launched too late, and so many other stereotypical startup world statements, I made all the mistakes to realise why those statements are so common. I'd do it all over again, to be honest. It has been an adventure of a lifetime, so to say.
And it worked out well (again, because I persevered, I looked to it that I learned from each failure and made the most of each situation I found myself in, no matter how bad).
After Crestify, I made more products. I made KindleClipper. Never launched it. KTool launched near the same time and gained decent traction so the idea was never the problem. I then made Solvemigo, ChatGPT on Telegram. I did launch it, so yay! I got dozens of upvotes on Product Hunt, never made a single penny. I did write about it Lessons Learned From Failing My First Product Launch and Charging ahead vs moving on from a project.
I then made PenPersona, inspired by Audio Pen. I fell into the tweaking and tinkering trap again. I chose to make it in Rails with all the turbo stuff. I got more interested in doing fancy stuff with the tech stack and lost sight of the bigger picture. Again, the premise of PenPersona is one I believe in even today. AI still writes like a robot, AI writing still lacks character and still doesn't match my own writing style. But I kept on tinkering and tweaking, I made a fancy pants glassmorphic UI. Never launched.
Then I took a step back. I knew something wasn't working and I needed to change or else I'd keep falling short of my potential, of my goals.
I came across the small bets approach. I learned to fight my perfectionism. And I made a tiny product. Maker Resources. A simple list of resources and tools for product makers, that had no scope for me to code anything for, so no room to procrastinate. I set a goal of making $10, and a stretch goal of $100. $10 because that's a no-brainer amount. I wanted to remove all hesitation someone would have in pressing buy now. I made $10 within 30 minutes of launching on Product Hunt, and I never marketed it beyond that but after a few months, maybe 6 or 9, I crossed the stretch goal too. In order to fight my perfectionism, I went back into my landing page, and added typos in. I hate typos and odd grammar so it was pretty difficult to purposely add those things into a landing page I was going to share with the world. Hence I did it. (Also decided to buy a domain name with a dash in it)
My most successful product to date is MemoryPlugin. I made it and launched it and got my first sale of it, all over the course of a weekend. I never intended to make a big product out of it, it was tiny in scope at launch, a niche extension for a niche product. A niche (people who want AI long term memory with TypingMind) within a niche (use TypingMind), within a niche (Use AI in general). Obviously now the product has matured a lot. It has more than 500 paying users, it has crossed $10K in revenues (I'm not going to speak as to exact revenues yet). When I launched MemoryPlugin to be honest I did not expect it would catch on, I did not expect anything. But small bets had also taught me, if you keep your expectations small, a product you hope to make $10 with looks very different than one you hope to make $100K with. I launched MemoryPlugin at $12.5 per year. A cheaper SaaS, I'm not aware of. I spent a lot of time marketing MemoryPlugin, often more than time spent on development. I released early and often. I avoided aiming to make things perfect. I avoided a lot of conventional prodyuct wisdom. People are still buying it, using it, and loving it.
If you are looking for a conclusion, I don't really have one yet. I just wanted to share this story, maybe something in here resonates with you and helps you make your own product / on your own journey.