How to find app ideas from App Store and Google Play reviews
· 8 min read
The fastest way to find app ideas from app reviews is to read the 1- to 3-star reviews of apps people already pay for, write down the job each reviewer couldn't get done, and look for jobs that many different people mention across several apps. Those clusters are gaps: demand that exists and that current apps fail to meet.
Most idea lists start from what someone thinks would be cool. Reviews start from what someone already tried to do and couldn't. That difference is the whole point. Below is the method, a real example from this week's data and the checks that separate a gap from a bad day.
In short. Apple processed over 1.3 billion ratings and reviews in 2025. The low-star ones are free customer research. Read recent 1–3 star reviews of paid apps, rewrite each complaint as a missing job, group the jobs, and build for the group with the most different people, apps and mentions of paying.
Why are app reviews a good place to find app ideas?
In 2025, Apple processed over 1.3 billion ratings and reviews, according to Apple's newsroom post The App Store stopped over US$2.2 billion in potentially fraudulent transactions in 2025. That's a huge, public record of people describing what they wanted an app to do. Unlike a survey, nobody prompted them. They wrote it because the app let them down.
Reviews have three properties idea brainstorming doesn't. They come from people who already looked for a solution, so the need is real. They name the app that failed, so you know the competition. And they're often specific: the price, the step, the moment it broke. Isn't that exactly what you'd ask in a customer interview?
The catch is volume and noise. Most negative reviews are about a single bug, a login problem or one bad update. The skill is filtering for the small share that describe an unmet need, then confirming that many people share it.
Which app reviews should you read?
Read 1- to 3-star reviews. A 2014 IEEE Software study by Khalid, Shihab, Nagappan and Hassan, What do mobile app users complain about?, coded 6,390 low-rated reviews of 20 iOS apps into 12 complaint types. Functional errors, feature requests and crashes were the most common. Hidden costs and privacy concerns hurt ratings most.
Those types are not equally useful for finding ideas:
- Feature requests are the obvious ones. "I just want to…" and "why can't I…" are people describing a product that doesn't exist yet.
- Hidden costs are often the best. They mean people wanted the thing enough to try it, then balked at the price model, not the product. That's a pricing gap you can build into.
- Functional errors and crashes are usually noise for idea hunting. They tell the existing developer what to fix. They only become an opportunity when the same job breaks across many apps for a long time.
Two more filters help. Read only the last 90 days, so you see the apps as they are now. And favour paid or high-grossing apps: people complaining about a weekly subscription have already shown the problem is worth money to them.
How to find app ideas from app reviews, step by step
The method fits on an index card. The work is in steps 3 and 4, where most people skip straight from reading reviews to building. A single review gives you an anecdote; a group of complaints from different people gives you evidence.
- Pick apps people already pay for. Start from a category's top-grossing chart, not the free chart. Reviews of apps people pay for tell you the problem is worth money.
- Read only the 1- to 3-star reviews from the last 90 days. Sort by most recent and filter to low ratings. Recent reviews describe the app as it is now, not three redesigns ago.
- Write each complaint as a missing job. Turn 'this app is a scam' into the job the person couldn't get done, such as 'send one fax without a weekly subscription'.
- Group complaints that describe the same job. Put complaints from different people and different apps that describe the same missing job into one group. One group is one candidate idea.
- Score each group on evidence. Count how many different people complain, whether they mention paying or switching, how recent the complaints are and how many existing apps they name.
- Check the original reviews before you build. Reread the best group's reviews at the source. If the quotes still describe one clear job that nobody does well, you have an idea worth testing.
Rewriting complaints as jobs (step 3) is the step that matters most. "Scam app, deleted" tells you nothing. "Wanted to send one fax, got asked for a weekly subscription" tells you the job, the blocker and the price model the person expected. Jobs from different apps start to line up once they're written this way, and that's how you see a gap.
What does a real gap look like?
When we checked on 14 October 2026, one of the highest-scoring opportunities on NicheGaps was an affordable pay-per-use faxing service: 36 complaints from App Store reviews of fax apps, all describing the same job. People need to send one fax, occasionally, and the apps they find want a weekly subscription.
Two of those reviews, both 1-star, both from summer 2026:
"Definitely NOT FREE as advertised. Tried to send first fax and I received the list of payment options. oh well…"
"I only tried to send one fax and it tried to charge me $10 … I can go to Staples and pay way less than that."
Read the second one again. The reviewer isn't refusing to pay. They're telling you the price they'd accept and where they'll go instead. That's a willingness-to-pay signal hiding inside a 1-star review, and it's the most useful kind of complaint there is. The gap isn't "a free fax app". It's a fax app priced per fax.
The same pattern shows up next door. A fair-priced document scanning opportunity groups 84 complaints across App Store and Google Play, like this one: "beautifully scanned all my documents only to let me know i need to pay to save them." Different apps, different people, same job: scan once, export once, pay a fair price for it.
How do you tell a real gap from noise?
Most complaint clusters fail at least one of these checks. Run all of them before you spend a weekend building. The question each one answers is simple: is this a problem many people have, that they'd pay to solve, that current apps don't solve?
| Check | Real gap | Probably noise |
|---|---|---|
| Who complains | Many different people | A few people, many reviews each |
| Where | Several apps, ideally both stores | One app |
| When | Steady over months | A spike after one update |
| Money | Mentions prices, refunds or switching | "Should be free" |
| The job | One job you can state in a sentence | "The app is bad" |
Competition deserves a note. A crowded category isn't a reason to walk away. The fax and scanner gaps above both sit in crowded categories. The complaints exist because the crowded apps all chose the same subscription model. A crowded field with one shared weakness is often easier to enter than an empty one.
Doing it by hand, or with a tool
By hand, this method works. Budget a few evenings per category: reading a few hundred reviews, rewriting them as jobs and grouping them in a spreadsheet. It's slow, but you'll understand the customer better than anyone who skipped it.
NicheGaps does the reading and grouping continuously, across App Store and Google Play reviews of top-charting apps, Reddit and Hacker News. Each opportunity is a group of complaints from different people, scored 0–100 on demand, competition, momentum, willingness to pay, urgency and breadth, with every quote linked to the review it came from. It's a starting point, not a verdict: the last step is still yours. Open the original reviews and decide whether you believe the gap.
Sources
- Apple Newsroom, "The App Store stopped over US$2.2 billion in potentially fraudulent transactions in 2025", 20 May 2026, retrieved 2026-10-14, apple.com
- Khalid, Shihab, Nagappan and Hassan, "What do mobile app users complain about?", IEEE Software 32(3), 2014, retrieved 2026-10-14, swag.uwaterloo.ca
- NicheGaps opportunity data and App Store reviews quoted above, retrieved 2026-10-14; each quote links to its source on the opportunity page.
Questions people ask
- Can you really find app ideas from app reviews?
- Yes. Low-star reviews are people describing, in their own words, a job an app failed to do for them. One review is an anecdote. Many reviews from different people, across several apps, describing the same failure, is evidence of demand that existing products don't meet.
- Which reviews are most useful for app ideas?
- Recent 1- to 3-star reviews of apps people pay for. Five-star reviews say what already works. One-star reviews about bugs in a single app point to a fix, not a product. The useful ones describe a job the person still needs done, and ideally what they'd pay or where they went instead.
- How many complaints do you need before an idea is worth testing?
- There's no fixed number, but count people, not reviews, and count apps. A dozen people across three apps describing the same missing job is a stronger signal than fifty angry reviews of one app's latest update, which usually just means that update broke something.
- Are Google Play reviews as useful as App Store reviews?
- Yes, and reading both helps. The same app often draws different complaints on each store because the audiences and devices differ. A complaint that shows up on both is less likely to be a platform quirk and more likely to be a real gap.
- Is it allowed to use app store reviews for research?
- Reading public reviews to understand what people need is ordinary market research. Don't republish reviewers' names, don't scrape faster than a site allows, and link back to the source when you quote someone. NicheGaps never shows reviewers' names and links every quote to the original.