TL;DR — The hardest part of building solo isn't execution, it's signal. Three passes turn a flood of newly launched startup projects into a shortlist worth a solo founder's attention:
- Watch what just launched. A live project is the market voting with real effort. Read it while the noise is still low.
- Read a lane's density before you enter it. Crowding confirms demand — it tells you to hunt for a seam, not to walk away.
- Judge replication difficulty. An opportunity can be real and still be wrong for one person.
Run the three in order and "what should I build?" turns from a guessing game into a weekly habit.
Who this is for, and where it stops
This is written for solo founders and independent makers who can ship but haven't picked a direction — people with more build capacity than conviction. Most of them don't lack ideas; they lack signal about which ideas are real, early, and small enough for one person to carry.
This playbook covers discovery: surfacing candidates and cutting them down to a shortlist. It stops there. Validating demand, pricing, and launch are separate jobs — the competitor research playbook picks up where this one ends.
The three passes at a glance
Each pass throws away more than it keeps. That's the point: the value isn't in seeing everything, it's in what survives all three.
a flood of new projects goes live
|
[1] just launched?
|
[2] lane crowded?
|
[3] can one person ship it?
v
a short list worth your week
How to read the examples below. They aren't picked for traffic or press coverage. They're recent, still forming, and chosen because the work behind them looks like something one person could realistically carry — the same filter you'd apply yourself, just applied earlier and across far more ground than a single browser tab can cover.
Path 1 — Watch what just launched
You don't need to predict the future. New startup projects go live constantly, and each one is somebody betting real effort on a real problem. The edge is timing: read the signal while the noise is still low and before the lane fills in.
The hard part was never finding projects — far more go live than any one person could track by hand, and almost all of them blur together. What helps is the smaller set that rises to the top once you weigh how realistically one person could actually build it. The projects below come from that set: early, still forming, and already showing the shape of something a solo founder can study.
A centralized dashboard designed to help creators organize and track progress across multiple side projects.
Sidespace is a dashboard that helps creators track progress across multiple side projects. Launched recently and carrying a low replication difficulty — shorthand for how little team, capital, or specialist access it would take to rebuild — it is exactly the shape of a solo-friendly opportunity: a clear job, a small surface area, no moat you'd need a team to defend.
A privacy-focused tool for importing and archiving conversations from ChatGPT, Claude, DeepSeek, and Grok into a local, searchable database.
AI Chat Importer is a privacy-focused tool that archives conversations from ChatGPT, Claude, DeepSeek, and Grok into a local, searchable database. Same pattern — a narrow job, launched early, low friction to replicate. Note what it doesn't need: no marketplace, no supply side, no regulated data. That absence is the signal.
Path 2 — Read a lane's density before you enter it
Not every lane is worth entering. Once you group new projects by what they do, a few lanes clearly draw far more attention than the rest — and that concentration is information.
Right now the heaviest traffic sits in the AI and SaaS spaces, with marketing, developer, and finance tools close behind:
AI / SaaS ██████████
Marketing ██████
Developer █████
Finance ████
(relative weight, not counts)
The common advice — avoid crowded niches — is backwards for a solo founder. An empty lane usually means nobody has found a buyer there yet, and proving a market from scratch is the one job a single person is worst equipped to do. A crowded lane has already answered the demand question for you. What it changes is where you enter: pick a seam the incumbents are too broad to serve, not a head-on rebuild.
Browse the full picture by niche:
Path 3 — Use replication difficulty to judge if it's solo-friendly
An opportunity can be hot and still be wrong for one person. If shipping it needs a team, capital, or a license you don't have, it's a trap no matter how validated the market looks.
That's why each project carries a replication-difficulty read:
- Low — one person can ship a weekend prototype
- Medium — doable solo, but slower; expect weeks, not days
- High — needs a resource moat: a team, funding, licensing, or proprietary supply
The projects that land in the low tier are the ones actually within reach of a solo founder — worth a closer look before anything else. The card below is a reference point from a crowded lane: an AI tool at medium difficulty, useful for calibrating what "harder" looks like.
An AI-powered platform for translating, subtitling, and dubbing video content from YouTube, TikTok, and X.
The tier is a starting filter, not a verdict. A medium-difficulty project can still be right if you already have the skill it demands — the point is to know which constraint you're accepting before you commit a month to it.
Where this filter misleads you
Three passes are a shortlist machine, not an oracle. The failure modes are worth knowing before you lean on it:
- Early looks the same as abandoned. A project that launched recently and then went quiet reads identically to one that's still forming. Give a candidate a second look a few weeks later before concluding anything from its first appearance.
- Low difficulty cuts both ways. If one person can rebuild it in a weekend, so can everyone else who noticed. Low difficulty tells you the door is open; it says nothing about how long it stays open. Pair it with a lane where you have an unfair angle — an audience, a domain you know, a distribution channel already warm.
- The filter only sees what ships. Opportunities that live inside companies, in offline workflows, or in communities that never register a domain are invisible here. This is a strong signal for software, and a blind spot for everything else.
Treat a shortlist as the beginning of research, not the end of it.
Putting it into practice with GrowthRadar
You can do the first two paths by hand, but they're faster as a repeatable workflow:
- Open the homepage and filter by category.
- Scan the replication difficulty badge — shortlist the Low ones.
- Open a project's detail page and read its Solo Founder Angle.
Thirty minutes a week is enough. The pattern is simple: let the noise settle, see which projects rise, and follow the low-difficulty threads.
FAQ
How do I know which new projects are worth my time?
Recency plus reach. A project that went live recently still has a readable story — you can see what the founder chose to build first. Reach is the second half: if rebuilding it would need a team or capital you don't have, the insight is interesting but not actionable. Worth your time means both, at once.
What makes a project solo-friendly?
A narrow job, a small surface area, and no dependency you can't supply yourself. Solo-friendly projects tend to avoid marketplaces (two sides to bootstrap), regulated data, and anything needing enterprise sales. The shortest test: could you describe the whole product in one sentence, and does that sentence name exactly one job?
Isn't a crowded niche a reason to walk away?
Usually the opposite. Crowding means somebody has already proven people pay for the outcome, which removes the single riskiest unknown. Treat density as a signal to narrow — serve a segment the broad players ignore — rather than a signal to leave. The lanes to avoid are the ones that are quiet for a reason.
How often should this be done?
Weekly is enough, and more often adds little. New projects surface continuously, but the ones worth studying need a little time to show whether they're still alive and still moving. A weekly pass catches them early without reacting to every blip.
What if I already have an idea?
Run it through pass three first. If your idea sits in the high tier, you've learned something cheap and early. If it's low or medium, use pass two to see who else is in that lane — that's the fastest way to find out whether you're early or late, and it's the input the competitor research playbook needs.
Where to go next
Back to the question this started with: what should I build? The answer isn't a flash of insight — it's a filter you run on repeat until something clears all three passes. Recent, crowded enough to prove demand, small enough for one person.
Found a candidate worth validating? The next step is checking who else is already in the lane and how well they're serving it — that's researching its competitors.