TL;DR
Seek help when you are stuck
Always reflect on your actions. It's easy to read "learn from your users" or "iterative development" etc. But it's very hard to put those into practice if you are not focusing on your goal
Read, Read, and Read. Read biographies, read YC writings. Read tweets from successful builders, but
Stop following others blindly. Listen and learn from them. But RUN YOUR OWN RACE
I've been somewhat absent from the public for a while. After the last ProductHunt launch "failed," I wanted to return to the drawing board and carefully plan out localportal roadmap and actually decide if I should continue building it or change my focus to something new. Short "about me" incoming. I've been in the AI space since 2016 (I'm disappointed in myself for not positioning myself to catch the falling apple during the AI hype) and have been working on AI infrastructure since 2019. I did this for top firms in the world and made a good amount of money that I can survive without a job for 3 years, but at one point, I left. Starting my own company or building a solopreneur business wasn't the plan when I left, but localportal was an idea I had even before that. It's not turning out exactly as I initially envisioned, but it's taking shape. Localportal was a scratch for my own itch. One of the issue that was annoying to me was the easy access to the GPUs I have at my office or at my home and the ability to collaborate/share the resources with others. We'd always have some sort reverse proxy SSH'ing into it, sharing ssh keys with others. When somebody left, everyone had to change the key... it was a mess.
Three months after leaving my job last year, I discovered this thing called building in public and indie hacking (I originally thought indie hackers were hackers from India. What a joker). So that's how localportal was born. However, as an engineer, I made the typical mistakes an engineer would make while building the product.
Mistakes I made
Built in steps, not in iteration: I immediately started coding and made the first version - which was just an alternative to ngrok but only for http/https. It wasn't quite what I wanted, but I was proud that I had been building iteratively. However, my iterations weren't small enough, and I didn't validate what each iteration brought. I just built and built - so it's not really an iteration if that's what you're doing; it's more like steps. Steps are good, but iteration is what you want in the early days, realizing that now.
Didn't try to acquire users manually and quickly. I assumed the first set of users would automatically come when I announced on ProductHunt or Twitter.
Built fast but didn't talk about it anywhere, not even to my existing users.
Overengineered some parts of the product (I am still proud of what came out of it, though).
I tried to focus on making money or selling rather than solving a problem that I knew existed, a.k.a at some point, I forgot to scratch that itch I had.
I tried to follow what worked for others and moved away from who I really am. Like posting on twitter about random things that I don't care about just to build audience.
Localportal launched on ProductHunt on December 2nd. By that time, it had already been in active development for four months without any marketing. Localportal had a solid engineering foundation, but it wasn't very useful, nor did it offer any reason for users to choose it over other tools like ngrok. To be honest, I didn't expect the launch to be a grand success (I had been listening to builders for a few months and knew that not everything was gold and diamonds by then). However, I was disappointed that the ProductHunt launch was a complete failure. I didn't get a single valuable signup from the launch, though I did get some from tweeting about it on Twitter. Afterward, I stopped development for a week or so, took some time to reflect on what I was doing, and realized I needed help. I reached out to people I knew had experience in this area - the best decision I made after leaving my job.
Why I made localportal in the first place
Localportal is not a simple product. It should never have been the first product anyone would create when they start. To shape it into something useful, one would need at least 4 months of very focused work. However, I still have my reasoning, and I believe I made the right decision, at least in this case. Localportal is a necessary tool (I can explain this) for an AI developer if you are working with models and experimenting with them. If you are using an existing API, this is not for you. I wanted to be the former, and I believe the world will have many such developers, making localportal invaluable for all those people. My belief was that if I had the essential tool that would 10x my productivity when I worked on the actual thing that mattered, I would be, well, 10x more productive, and while doing that, I could also sell this to other people and make them 10x more productive.
What have I been doing? I already had an ideal customer profile in mind: me. I have had issues that I have been trying to fix. So,
I took products that had some overlap with what I built and tried to find the distinguishing factor.
I listed out other possible ICPs to see if I had a backup plan if my current ICP didn't work out (turned out there are many). The goal was to try to find a deep niche (not a narrow one).
Admitted that I am not an entrepreneur (yet) and that there is a long way to go.
Talked to existing users or people I thought would be able to use localportal.
It's changing
As I mentioned, I received help. They made me realize that I wasn't building for my audience, and I hadn't considered positioning my product, the landing page, or any part of the onboarding flow to make it appealing to a new visitor. I never tried to teach my audience what problems Localportal solves. Most importantly, I didn't prioritize tasks correctly (remember, I was just building). In a nutshell, my perspective towards my product needed to change. I needed to start thinking like an entrepreneur rather than an engineer. Sounds simple, right? Not so much when actually doing things. I am still conducting weekly calls with the circle I am in. We constantly examine the tasks at hand and reanalyze the priorities. (I'll talk about them soon, but for now, I want this write-up to focus more on what happened than how/who). As an entrepreneur, your priorities change.
If I were to give you the key takeways, these are those
Seek help when you are stuck
Always reflect on your actions. It's easy to read "learn from your users" or "iterative development" etc. But it's very hard to put those into practice if you are not focusing on your goal
Read, Read, and Read. Read biographies, read YC writings. Read tweets from successful builders, but
Stop following others blindly. Listen and learn from them. RUN YOUR OWN RACE
Perfect results count, not perfect process