ReviewFold
Guide · From reviews to ideas
How to find app ideas worth testing in app reviews
You do not need another list of app ideas. You need a problem described clearly enough to investigate. Reviews help because they put someone’s expectation next to the product they actually tried — and the gap between the two is where a question starts.
01 · The starting point
Start with a task someone already tries to finish.
A feature list is easy to write and hard to test. A task gives you something to watch for in the next review you read.
Choose an app that people use for a task you understand. Start from the task, not from a plan to copy every feature of that app — the features are one product’s answer, and the task is what the answer was for.
Read complaints and positive accounts together. A complaint shows you friction; a positive account shows you what a replacement would have to preserve, which is the part that usually gets left out of an idea list. Neither one works alone.
A rating is not a description
A one-star rating alone does not establish what happened. It does not say what they were doing, or what they expected instead. Read the sentence before you decide the review is useful.
Write down three things for each review that survives that test: the activity, the situation, and the outcome the reviewer describes. If a review only says “terrible,” keep it as context until you find something more specific.
02 · The clue
Three reviews, one shared detail.
A historical sample of 200 public Threads reviews from the US App Store, collected on September 23, 2026. The three selected reviews below mention the same device.
“Does not work properly on iPads.”
“plz adapt iPad ui”
“virtually useless on the iPad I use all the time”
The shared detail is the device. Their descriptions differ: one reports that the app does not work properly, one asks for iPad UI adaptation, and one contrasts the experience on an iPhone with an iPad they use regularly.
They do not describe identical symptoms, and they do not tell us whether the difficulty involved reading, composing, navigation or something else. That gap is not a flaw in the evidence — it is the shape of what you still have to find out.
03 · The question
Turn the clue into something you can answer.
A question you can investigate is narrower than an idea, and far more useful than a market claim.
The question these three reviews support
What are frequent iPad users trying to do, and where does the experience get in their way?
Notice what the question does not do. It does not name a product, a feature or a business. It names a group of users, the situation they are in, and the thing you do not yet know — which is what makes it answerable by a conversation or a walkthrough rather than a guess.
It also does not yet give anyone a reason to build a new social network. The answer might point to a layout issue in an existing app, a workaround nobody has written down, a narrower tool that does one part well, or no remaining problem at all. Each outcome suggests a different next step.
This is a manual reading of selected historical evidence, not a test of the current app. ReviewFold is not affiliated with or endorsed by Threads or Meta.
04 · The brief
Write the brief while parts of it are still blank.
Seven fields. Four have a draft entry; three need further checking. Draft entries include hypotheses and planned questions, not just established facts. The open ones are the reason to do research before building anything.
4 written now3 still open
- Possible audience
- People who use an iPad regularly and want to use Threads there.
- Task
- Still needs clarification: reading, composing, or another activity Open
- Evidence
- Three selected reviews mention the iPad experience. These are selected accounts, not an exhaustive count of iPad feedback.
- Existing choice
- Threads. Investigate the current app and browser experiences before proposing a replacement.
- What a new option must preserve
- Unknown. Ask what people value about their current workflow. Open
- Next question
- Ask a relevant user to walk through their most recent attempt and show where they got stuck.
- Reason to stop or change direction
- The current experience already works, the difficulty is rare, or an existing workaround is good enough. Open
The unfinished fields are not a gap in the page. They show what to learn next. The stop conditions are proposed in advance; whether any of them holds remains open.
05 · Existing options
Check what already exists before proposing a replacement.
Try the relevant current workflow wherever you can legitimately access it, read the product documentation, and read recent feedback. Record what you checked and when — a replacement proposed against a two-year-old memory of the app is a guess wearing a plan’s clothes.
If you are looking at several apps, keep their evidence separate. Similar labels hide different problems: a request for a larger interface is not the same as a broken posting flow, and merging them produces a requirement nobody actually asked for.
Check feasibility at the same time. A companion app may depend on data access or platform permissions you do not have. A convincing complaint does not remove those constraints, and no amount of reading reviews will.
06 · First-hand evidence
Ask about the last real attempt.
Take the question to people who actually do the task. Ask about a specific occasion rather than the general habit.
- What were you trying to finish the last time this happened?An ordinary occasion, not the worst one they remember.
- What did you do instead when it did not work?A workaround is a competitor you have not noticed yet.
- How often does it come up?Frequency helps you assess how often the problem matters.
- What works well enough that you would not want to lose it?The list a replacement has to preserve before it can win.
- What would make trying something else worthwhile?The cost of switching, in their words rather than yours.
Avoid opening with “Would you use my app?” A description of what someone did last Tuesday gives you more to investigate than a polite reaction to a product that does not exist yet.
07 · The smallest test
Test the uncertain part, not the whole idea.
Once the task is clearer, choose one important uncertainty to test. Demand, switching costs and feasibility may still be unresolved.
That might be a sketch of a different layout, a walkthrough of a proposed flow, or a small prototype of the one step you are unsure about. It does not need to be a product, and it should be small enough to learn from before you commit to building.
Before you run it, write down what result would make you continue, change direction or stop. You are looking for evidence that the alternative helps with the task, not agreement that the idea sounds interesting — those are different reactions, and neither alone establishes a business opportunity.
Willingness to switch and willingness to pay still need their own evidence. Reviews can help you choose whom to talk to and what to ask; they cannot settle either question.
Limits
Reviews narrow the question. They do not answer it.
Knowing where a brief stops is what keeps you from presenting a guess as research.
- They cannot size a marketThree selected reviews are three public accounts. The sample is not the user base.
- They cannot price a switchWanting a better iPad experience is not the same as paying for one. That needs its own evidence.
- They cannot prove feasibilityPlatform rules, data access and permissions decide what can be built, and they are read somewhere else.
- They cannot replace a conversationA conversation or walkthrough can clarify context that a short review leaves out.
Current support, stated plainly
- Apple App Store: US and Canada. Google Play: US.
- Up to 200 recent public reviews per report.
- One app and one region per report; no cross-app comparison.
- A free preview shows themes and counts; the paid report adds review excerpts, PDF and CSV.
- No opportunity scoring, market sizing or automated idea generation.
Start with the evidence
Assemble the accounts, then decide what to build.
ReviewFold groups related feedback from a recent public-review sample of one app and one region. Use it to find candidate patterns and to get back to the original reviews. The brief, the comparison with other options and the decision to build stay with you.