So it's day 1 of the $0 to $50k MRR in 18 months challenge and more often than not the hardest part is starting.
You know the deal. You've got a blank piece of paper in front of you, it's looking you deadpan in the eye, taunting you with all the ways you could start but providing no answers, which sends your mind racing as you talk yourself out of it.
And that's exactly where I landed last night on Day 0. I got back from the office, ran around the garden with my daughter as she played on her new hot pink slide (which she loves by the way), sat down for dinner as a family, discussed the days events and then I said good night to my wife and opened my laptop.
The blank page was staring at me, I could already feel the doubt in my mind.
And when that doubt creeps in, I like to start giving myself constraints, because I could sit and plan all day and get no where, I've made that mistake far too many times in my life. Over planning, leads to over thinking which leads to zero action.
And action provides clarity as long as we can get over the pain of uncertainty.
So I started thinking about niches. In fact I started thinking about niches, of niches of niches. Maybe I niched too far, but I want to get specific.
I'm not going to be building B2C apps, that often requires audience (which I don't have at this stage), it often requires serious VC capital (which I don't have, nor do I want) and it often requires huge adoption for the numbers to work out well. You literally burn money to win market share.
At the end of the day, I'm not here to compete with app stores or build a massive VC funded SaaS, I want to launch a series of profitable micro products.
I'm here to solve an incredibly painful and specific problem for a hyper specific person.
You see, constraints provide clarity. So what are those starting constraints?
- b2b
- hyper specific problem
- hyper specific person
Getting the ball rolling
Now I had a rough idea of what and who I was going after, so naturally the next question was, what are those businesses, how specialised are they and could I derive painful problems looking from the outside in?
So I created a new project, pointed Claude Code at it and got to putting a research plan together.
Within that folder were two files providing Claude with the mission context, once he had parsed and asked "Ready when you are — what would you like to do with these?"
I dropped this prompt.
So we need a plan and we need research. Lets start with research, and lets think about Niching into the niche of a niche, each app targets a hyper specific avatar. For example lets think about Dr's.
General Practitioner -> Haematologist -> Sickle Cell Specialist -> Sickle Cell Transition Specialist.
We want to target and conduct reaerach within the most niched pool, specifically businesses with money to spend, a problem we can identify and a decent sized pool of individuals.
I want to explore those niches first and foremost, just a rough overview because for each micro product I ship, I'll pick a hyper specific niche with a specific problem so we need to test the validity of them.
I used Fable 5 on Ultra, we spun up 11 subagents and sent them out on a serious research quest.
Now we also need to ensure that the agents stayed honest so for phase 2 we needed to rank the opportunities according to my constraints and available time
So a sceptical judge dedupes everything and scores each candidate 1–5 on seven dimensions: pain severity, buyer budget, TAM fit (sweet spot ~5k–200k businesses), reachability via paid demand, solo-buildability in ~20h/week, competition gap, and founder fit.
The goal was to return a ranked top 12–15 shortlist plus a "notable rejects" list, shiny-looking niches that fail a filter, documented so we never re-litigate them.
Once that shortlist landed, I picked out 4 potential concepts and instructed Claude to run a validation test across items like cost, feasibility of concept, legality, ease of solving the pain and tying that to the revenue path etc.
Four avatars and specialised businesses entered the validation test and one came out on top with a Green Go Flag.
I still have another 11 to check and run, but the main thing was I went from a blank page to having real avatars, with potentially real problems and now I just needed to figure out how easy or hard solving those problems via software was.
So that's what I'll be working on this evening, running a handful more through validation and exploring the feasibility a little deeper on that one avatar and business, either way, I'll be building something by this weekend.
Catch you in the next one.
Tariq