What Makes a Good App Opportunity (And How We Score Them)

A bad app is not the same thing as a good opportunity.

There's a flashlight app with a 1.2-star rating. Terrible, right? Sure. But there are also about 400 other flashlight apps, several of them rated 4.8 stars. The problem is already solved. Nobody needs you to build flashlight app #401.

Then there's a music notation app with a 2.0-star rating, 249 reviews, charging $2.99, sitting at #42 on the paid charts, and it hasn't been updated in eight years. The reviews are full of musicians saying "this is the only app that does this and it barely works." There are zero well-rated alternatives.

Both are bad apps. Only one is a real opportunity. The difference comes down to five things.

Signal 1: Demand that you can actually measure

The first question is always: does anyone actually use this thing?

Rating count is the best proxy we have. An app with 5,000+ ratings has serious usage behind it. Each rating represents way more users who never bothered to rate. An app with 50 ratings? That's a niche inside a niche inside a niche.

Chart position is even better when it's available. If an app is on Apple's Top Charts despite having a terrible rating, that's demand you can't argue with. People are downloading it right now, knowing it's bad, because they need what it does and can't find anything better. We found 301 apps in exactly that situation.

Category size matters too. An opportunity in Education or Finance is a bigger pond than an opportunity in, say, Catalogs. More potential users, more room to grow, more revenue ceiling.

Signal 2: Frustration that's specific and fixable

A low star rating by itself doesn't tell you much. You need to know why people are angry. And more importantly, whether the thing they're angry about is something you can actually fix.

Some complaints are gold:

• "Crashes every time I open it on iOS 18" (fixable by simply building with current tools)
• "The UI is impossible to figure out" (fixable with basic design sense)
• "I paid $4.99 and now they want a subscription too" (fixable with an honest pricing model)
• "No dark mode, no widget, no Apple Watch support" (fixable, and you get a feature list for free)

Some complaints are worthless:

• "I don't like the concept of this app" (not a product problem)
• "Too many ads" on a free app (that's... how free apps work)
• "This should be free" on a $0.99 app (these people won't pay you either)

We read through 485,929 reviews across the dataset and extracted the top complaints for each app. The ones that score highest are the ones where the complaints are specific, recurring, and clearly fixable by a competent developer.

Signal 3: No good alternative exists

This is the one that trips people up. You find a bad app with tons of users and think "great, I'll build a better one!" But then you search the category and discover three 4.8-star alternatives already exist. The opportunity was real... two years ago. Someone already took it.

For every app in our dataset, we checked the competitive landscape. The question isn't "are there other apps in this category?" (there always are). The question is: "is there a well-rated app that does the same specific thing?"

When the answer is no? That's when it gets interesting. And honestly, we were surprised how often the answer is no. There are entire niches where every single app is rated below 3 stars. Nobody has bothered to build the good version yet.

Signal 4: Revenue evidence

"Will people pay for this?" is a scary question when you're brainstorming app ideas in the shower. It's a lot less scary when the answer is sitting right there in the App Store listing.

Paid apps with active users are the clearest signal. Someone is already spending $2.99 or $4.99 or $9.99 on this thing. The price point is validated. The willingness to pay is proven. You just need to build something worth paying for.

We estimate revenue for every paid app in the dataset based on price, rating velocity, and category benchmarks. It's not going to be exact down to the dollar, but it tells you the difference between "this makes coffee money" and "this is a real business." That distinction matters when you're deciding where to spend your weekend.

69 apps in the dataset are paid, abandoned, and still generating revenue. People are literally handing money to developers who stopped caring years ago.

Signal 5: Can a solo developer actually build this?

An opportunity that requires a 10-person team and $2M in funding isn't an opportunity for our audience. We specifically score for solo-dev feasibility.

The best opportunities tend to be apps that are conceptually simple but poorly executed. Niche utility apps. Specialty calculators. Reference tools. Education apps for specific subjects. These are things one person can rebuild in SwiftUI in a weekend or two, especially since the original was probably written in Objective-C back when Obama was president.

We flag build difficulty for every entry: low, medium, or high. Most of the top-scoring opportunities land in the low-to-medium range. That's not a coincidence. Simple apps are the ones most likely to be abandoned (the developer got bored) and most likely to be replaceable (the rebuild is straightforward).

How it all becomes a score

Every app in the dataset gets an opportunity score that combines all five signals into a single number. High demand, high frustration, no competition, proven revenue, and solo-dev feasibility all push the score up. The opposite pulls it down.

We then tier them:

The top 13 entries are our Top Picks. These are the ones where all five signals line up and the opportunity is hard to ignore.

A real example from the dataset

Here's an actual S-tier entry so you can see what this looks like in practice:

High demand (on the charts, real revenue), severe frustration (2.0 stars, crashes, paywalls), zero competition, and the build is hard but doable for someone with audio programming chops. Every signal lines up. That's what an S-tier opportunity looks like.

The dataset has 6,219 entries like this. Not all S-tier, obviously. But every single one scored, tiered, and ready to evaluate.