Goal
Touch the 5k MRR mark. That was the goal for my first product, because I wanted to apply to Y Combinator.
Strategy
I found the gap while researching what to build. The neurodivergent community was hitting friction with how configurable Notion and Obsidian are, and the pain points were sitting in the open on Reddit: they wanted something natural, engaging and easy for recording goals, habits and plans. So customisability was the friction, not the feature. I reduced friction while targeting only that one segment, iterating every feature and pain point in the Discord and Reddit communities. I also ran my own Discord server and asked it for feedback every two weeks.
Acquisition
Purely organic. I A/B tested the channels and TikTok won, so I posted there 3 times a day, promoted in Reddit and Discord every 2 days, and wrote a blog post every week.
Difficulties
The audience turned anti-AI over the potential pollution effects of data centres, and turned hostile to the exact features built for them. I spent months rebuilding it AI-free on continuous feedback, but stripped of its main features it never recovered. Payments fought back separately: Stripe first, but registering for VAT in every EU country was too much overhead for one person, so I moved to Paddle as a merchant of record. Paddle rejected us because they would not accept AI chat features at the time. Lemon Squeezy accepted, also a merchant of record.
Outcome
I killed it rather than keep rebuilding. The same community had recommended several of those AI features while I was building them, which is why the lesson I learnt is not to follow feedback blindly: an audience says one thing and does another. Some outcomes cannot be predicted, so the skill is knowing when to stop instead of fighting a shift you cannot reverse.