The hard part of being a solo developer is not building the app. It is choosing the right one to build. Learning how to find app ideas that are actually worth your months of work is the skill that decides whether you ship something people want or something only you wanted.
I build apps in public under softDev23, and I recently spent a serious block of time on exactly this question for my next projects, part of building in public as a solo developer. I am not going to hand you my specific shortlist, because that is the part I am actually building. But the method is more valuable than the list, and the method I am happy to share.
Deciding what to build next is easier once the ones you already have are under control, which is the problem I tackled in how I manage multiple app projects.
This is how I separate an idea that looks good from one that is genuinely worth it.
A good idea is not the same as a good bet
Most idea advice stops at finding a problem you have. That is a fine starting point, but it is only the first filter. Plenty of real problems are terrible bets because the market is saturated, impossible to reach, or owned by someone with a thousand times your resources.
So I stopped grading ideas on whether they were good and started grading them on whether they were a good bet for a one-person team. Those are different questions. A good bet is a real problem that also happens to be winnable by someone with no budget and no audience yet.
That reframe kills a lot of exciting ideas early, which is the point. The goal is not to feel inspired. It is to avoid spending six months on something I could never have won.
Start with demand you can actually see
The first thing I look for is evidence that people are already trying to solve this, in their own words. Not a hunch that they might want it. Proof that they are searching for it and talking about it now.
That means search demand and real conversations. Are people typing this into a search engine in meaningful numbers? Are they asking about it in forums and communities, complaining about the existing options, and asking what to use instead? If nobody is searching and nobody is talking, there is no demand to capture, no matter how clever the idea feels.
Paul Graham makes a related point in his essay on how to get startup ideas, that the best ideas come from noticing real, present needs rather than inventing ones. I treat visible demand as the price of entry. If I cannot see it, the idea does not advance.
Then check whether you can actually rank or get found
Demand with no way to reach it is a trap. The second filter is distribution. If I build this, can people who want it actually find it without a marketing budget?
For a web tool, that mostly comes down to search. I look at who already ranks for the relevant terms and how strong they are. If the first page is wall-to-wall giant sites, a new one-person project has little chance, and the idea drops. If the first page is weak, thin, or full of pages that do not really answer the query, that is an opening.
For an app, the same logic applies to the app store and to the blog posts that send people there. The question is never just whether there is demand. It is is there demand I can reach from a standing start. A market I cannot get found in is not my market, no matter how big it is.
Red-team your favorite idea on purpose
Once an idea survives demand and distribution, I do the thing that is most uncomfortable and most useful. I argue against it as hard as I can.
I ask who already owns this space, why they would crush a newcomer, and what would make this fail. Favorite ideas get a free pass in your head, which is exactly why they are dangerous. The ones that feel obviously great are often the ones you have not stress-tested, because you did not want to.
Half of my best decisions in this round were really decisions not to build something. An idea that looked perfect fell apart the moment I asked who else was already doing it well. Killing it on paper costs an afternoon. Killing it after six months of building would have cost a year.
Look for the underserved edges
The ideas that survive all of this tend to share a shape. They are specific, a little unglamorous, and sit in a gap that the big players cannot be bothered to fill.
Broad, obvious categories are where everyone competes, and nobody wins from zero. The winnable spots are usually narrower. A specific problem for a specific group of people, served better than the generic tools bother to. The narrowness that makes an idea feel small is often exactly what makes it winnable.
This is the same logic behind how I sort my own products in my post on the best apps for ADHD, by the specific problem each one solves rather than the broad category it sits in. Specific wins because specific is reachable.
Pick for the long game, not the dopamine
A last filter that took me a while to respect. Some ideas are exciting to start and miserable to sustain. Some are quietly motivating for years. As a solo developer, the second kind matters more, because the graveyard is full of clever apps whose makers got bored.
So I weigh whether I will still care about this in a year, and whether the work compounds. A tool that feeds my blog, that I will keep using myself, that gets a little more valuable with every post, is worth more than a flashier idea I will abandon by autumn. I went deep on the realities of finishing and shipping in my post on the numbers behind launching my first habit tracker, and stamina was a bigger factor than talent.
The research that produced my current shortlist used a loop, not a single sitting, which is the same approach I describe in my post on a reliable AI agent workflow. I kept tightening the list against real data until it stopped changing.
Why I share the method and not the list
You might notice I keep describing how I evaluate ideas without naming the ones I landed on. That is deliberate and part of the method, too.
The shortlist is the part with real value, because it is what I am actually going to build. Handing it out would just invite someone with more time to race me to it, and on a winnable idea, the gap between first and second can decide everything. The method, on the other hand, costs me nothing to share and helps you find your own ideas, which will be different from mine anyway.
This is a useful instinct for any solo builder. Be generous with how you think and protective of what you are currently betting on. The “how” builds your reputation and helps other people. The “what” is your edge, and edges are worth keeping until you have shipped.
It also keeps me honest. If I can explain the filter clearly enough for you to use it, I understand it well enough to trust my own results. A method I can teach is a method I have actually thought through, rather than a gut feeling I am dressing up after the fact.
A quick test you can run this week
You do not need fancy tools to start. You need an honest hour. Take your current favorite idea and run it through a fast version of the same filter.
First, search for it the way a frustrated user would, in plain words, and see what comes back. Are there real people asking about it, or just companies selling at them? Then look at who ranks on the first page and ask whether a new, small site could realistically join them. Then spend ten minutes trying to talk yourself out of the whole thing, listing every reason it might fail and who might already own it.
If the idea is still standing after that hour, it has earned a deeper look. If it fell apart, you just saved yourself months for the price of an afternoon. Either way, you learned something concrete, which beats sitting with a vague feeling that an idea is probably good.
The point of the quick test is not to be rigorous. It is to build the habit of checking before committing. Most bad bets are not bad because the idea was dumb. They are bad because nobody ran even the cheap checks before pouring in the expensive months.
How to find app ideas, in one line
If you take one thing from this, let it be the order. Demand first, distribution second, a hard red-team third, and your own stamina last. An idea has to clear all four to be worth your months.
That is really all how to find app ideas comes down to for a solo dev. Not a flash of genius, but a filter you run honestly, that is willing to kill the ideas you like most when the evidence says they will not win. The flash of genius is overrated. The filter is what ships.
