Niche positioning is the only product strategy that makes sense when you're one person competing against companies with hundreds of engineers. I don't try to build a better Bible app than YouVersion, a better recipe manager than the giants, or a general exam-prep platform. I build for one exam, one doctrinal tradition, one country's home cooks. The narrower the audience, the more the product feels like it was made for exactly them — because it was.
This isn't a romantic idea. It's the practical conclusion I keep reaching after shipping product after product solo. When you can't out-resource the incumbents, you out-specify them. You pick an audience so precise that a generalist tool can only ever serve them badly, and then you serve them exactly. Three of my products make the argument better than any abstraction, so let me use them.
One exam is a feature, not a limitation
Admitere Drept is an exam simulator for a single test: the admission exam for the Faculty of Law at the West University of Timișoara. Not law admissions in general. Not Romanian university entrance. One faculty, one city, one annual cohort of teenagers.
By any market-sizing logic, that's a reason not to build it. The total addressable market is brutally small and resets once a year. But that smallness is exactly what lets every feature be exact instead of generic. The countdown timer matches the real exam. The question categories — vocabulary, analogical reasoning, inductive reasoning — match the real syllabus. The final-average calculator uses the actual 50/50 weighting (Baccalaureate plus test) that UVT publishes. There's no "configure your exam type" wizard, no abstraction layer covering twelve universities I'd model badly. A generalist competitor literally cannot ship the average calculator, because the formula is specific to this one faculty.
That's the core mechanic of niche positioning: specificity that a broad product structurally can't match. The narrowness isn't a constraint I tolerate. It's the moat.
Language and tradition as a moat
Sometimes the wedge isn't a single use case but a market that the giants don't bother to serve well. Studiază Biblia is an AI Bible-study companion for Romanian believers. The English-language incumbents — YouVersion, Logos — are enormous and excellent. Competing with their feature lists head-on would be foolish.
So the product doesn't. It picks two axes the incumbents are weak on. The first is language: a modern, mobile-first, AI-powered study tool in Romanian, for a market dominated by English apps. The second is doctrine. The AI companion, Parakletos, is grounded explicitly in one theological tradition (Seventh-day Adventist), baked into its prompts and content rather than bolted on. A generic Bible app has to stay denominationally neutral to serve everyone, which means it serves no specific tradition deeply. Studiază Biblia does the opposite on purpose.
Language is an underrated moat. It's not just translated strings. It's the doctrinal vocabulary, the specific translations that matter to this audience (Cornilescu/VDC, Fidela), the cultural assumptions a Romanian believer brings that an English-first app never modeled. A multinational can localize a UI; it's far slower to localize judgment. That gap is where a solo founder lives.
Build for who you actually understand
Rețetele Mele is a recipe organizer for Romanian home cooks — paste a URL, the backend extracts the recipe, and your collection lives in one offline-first app. Recipe apps are a crowded category globally. The niche here isn't a clever mechanic; it's an audience I can picture concretely: Romanian-speaking people with recipes scattered across blogs and social video who want them in one searchable place, in their own language.
Everything points at that person. The UI, the prompts, the docs are all in Romanian. That focus makes a hundred small product decisions easy, because there's a specific user to decide for instead of a global average to compromise toward. Worth being honest, though: Rețetele Mele is still an MVP — working locally, not yet deployed, with production infrastructure deliberately deferred. The niche makes the product easy to define; it doesn't make it easy to finish.
The trade-off you're actually accepting
Niche positioning has a real cost, and pretending otherwise would be dishonest. The addressable market is small — sometimes brutally so. Admitere Drept's audience is a single cohort that shows up once a year; the admission calendar is the product's heartbeat. You cannot grow a product like that by widening it, because widening is exactly what destroys the specificity that made it good. The moment Admitere Drept tries to cover twelve faculties, it becomes a worse version of a generalist tool.
So the honest framing is this: a niche product is rarely a company on its own. It's the best tool for one job, and it earns its keep by being exactly that. This is why I run a portfolio instead of betting everything on one idea — it's the whole logic behind building 11 products as a solo founder. Several small, sharp products beat one product trying to be everything, when you're the only one shipping.
There's a second trade-off worth naming: outcome data is hard to get on a small audience. With Admitere Drept I can state the architecture is solid and the app is live on both stores, but I can't yet prove that students who drill on the simulator score better on the real exam. The cohort is small and the signal is slow. Niche means you trade volume for depth, and depth includes the painful kind — fewer users to learn from, slower feedback.
How monetization works when the audience is tiny
The obvious objection to niche is money: how do you make anything from an audience this small? The answer is that you stop optimizing for revenue scale and start optimizing for fit and friction.
Admitere Drept is freemium. The free tier is usable but ad-supported. Pro is a one-time purchase of 19.99 Lei — roughly four euros — that unlocks unlimited tests, full explanations, and removes ads. On a market this size, that pricing isn't revenue optimization. It's a low-friction way to let the students who are serious about the exam clear the ads and study without limits. The product earns trust by being genuinely useful for free, then asks for a small, fair amount from the people it's clearly helping.
Studiază Biblia takes a different shape: a free Romanian MVP now, architected so the same codebase can flip — by configuration — into a paid international tier across six languages later. The niche launch builds the product and the audience; the broader monetization is a deliberate later phase, not the opening move. That's a pattern I trust: prove the thing works for a specific group, keep the architecture honest about a wider future, and don't pretend the money is here before it is.
The unifying principle across all three: when the audience is small, every unit of trust matters more. You can't afford to feel like a generic app monetizing a captive user. You earn the upgrade by being so specific that paying feels like supporting a tool that was, visibly, built for them.
The bet
Niche positioning is a bet that depth beats breadth when you're outmatched on resources. Pick an audience a generalist can only serve badly. Make every feature exact for them. Use language, tradition, or a single use case as the moat a bigger company is too broad to cross. Accept the small market as the price of a product that actually fits, and monetize with fairness rather than scale.
It won't make any one of these a unicorn. That was never the goal. The goal is a set of products that each do one thing better than anyone bothering to compete, for people who can tell the difference. If you want the rest of the portfolio, the projects page is the map.