Your competitor's users are already writing your product roadmap. They do it in public, in the app stores, one short review at a time. Most teams read those reviews in the wrong way. They scroll, they pick a few angry ones, and they call it research. This article shows a more careful method. You read complaint clusters across two apps side by side. You look for the gap that one app has and the other does not. Then you check that gap against what the company says about itself in official documents.
This is not investment advice. It is a method for reading public text with some discipline.
Why a single app tells you very little
Every app has complaints. Log in fails. The update broke something. The subscription is too expensive. If you read one app on its own, you see a wall of problems and you cannot tell which ones matter. Some complaints are universal to the category. Others are specific to one product.
The useful signal sits in the difference. Say both apps get complaints about price. That tells you the category is price sensitive. It does not tell you where your competitor is weak. Now say one app gets steady complaints about sync between devices and the other app almost never does. That is a gap. It points to something the second team solved and the first team has not.
So the unit of analysis is not the review. It is the complaint cluster, compared across two products in the same category.
Step 1: pick a fair pair
The comparison only works if the two apps do the same job for the same kind of user. A budgeting app and a banking app both touch money, but their users complain about different things. Pick two products a real customer would choose between.
Match the platform too. iOS and Android reviews can differ because the apps themselves differ per platform. Compare iOS with iOS and Android with Android. If you want both, run them as two separate comparisons.
Match the time window. An app that shipped a large redesign last month will have a spike of complaints that says more about change than about quality. Use the same date range for both apps, and note any major release that falls inside it.
Step 2: collect the text
You do not need a paid tool to start. The free options work for a first pass:
- Read the public store pages directly. Both major stores show recent reviews and let you sort them. This is slow but free.
- If you own one of the two apps, your own developer console gives you your reviews in bulk. Apple has App Store Connect and Google has the Play Console. Neither gives you a competitor's reviews in bulk, but they give you a clean baseline for your own side.
- Paste what you read into a spreadsheet. One row per review, with the date, the star rating and the text.
A paid scraper saves time once you want hundreds or thousands of reviews per app. It does not change the method. Keep the raw text. You will want to go back to it.
Step 3: tag complaints, not stars
Star ratings are a blunt instrument. A three star review can hold a sharp, specific complaint. A one star review can be about a delivery driver who has nothing to do with the app. Ignore the stars for now and read the text.
Give each negative review one or more tags. Keep the tag list short and concrete. Good tags describe a part of the product: login, sync, notifications, billing, search, performance, support, data loss, ads. Bad tags describe a feeling: frustrating, bad, annoying. A feeling cannot be fixed. A part of the product can.
Tag both apps with the same list. If you invent a new tag halfway through, go back and apply it to the reviews you already read. Otherwise the counts will lean toward whichever app you read last.
Step 4: compare shares, not raw counts
A popular app gets more reviews than a small one. Raw counts will make the bigger app look worse at everything. So compare shares. For each app, divide the number of reviews with a given tag by the total number of negative reviews you tagged for that app.
Now line the two apps up per tag. Three patterns tend to show up:
- Both apps share a similar share for a tag. That is a category problem. It is worth knowing, but it is not a weak spot for either side.
- One app has a clearly higher share than the other. That is a candidate weak spot.
- A tag appears for one app and almost never for the other. That is the strongest kind of candidate. Something is broken or missing in one product and not in the other.
Be honest about sample size. If you tagged a few dozen reviews per app, a small difference in shares is noise. Only treat a gap as real when it stays visible after you read more reviews, or when it shows up again in a different month.
Step 5: read the cluster, not the count
Once you have a candidate, go back to the raw text for that tag. Read every review in the cluster. You are looking for three things.
First, specificity. Do users describe the same failure in the same words? "Sync deletes my entries when I switch phones" is a precise complaint. Many users saying that is a real product gap. Vague complaints scattered under one tag may be several different problems grouped together.
Second, persistence. Does the complaint show up across many weeks, or only in the days after one release? A short spike often gets fixed. A steady complaint that survives several releases points to something structural. That can be an architecture choice, a staffing gap, or a business decision the team has made on purpose.
Third, the workaround. Users often tell you what they did about it. "I switched to the other app for this" or "I export to a spreadsheet every week now" tells you how much the problem costs them. A complaint with a costly workaround is worth more than a complaint users shrug off.
Step 6: watch for fake and filtered reviews
Review text is not a clean dataset. Some reviews are fake, some are incentivized, and some platforms filter or reorder what you see. In the United States the Federal Trade Commission treats fake and deceptive reviews as a consumer protection issue and publishes guidance for businesses on endorsements and reviews. That tells you the problem is real enough for regulators to address it. It does not tell you how much of any given store page is affected.
A few practical checks help. Clusters of near identical wording posted in a short window are suspect. Reviews that praise one app by attacking the other by name deserve a second look. Treat these as noise and keep them out of your shares. If you cannot tell, leave them in and note the doubt.
Step 7: check the gap against what the company says
Reviews tell you what users feel. They do not tell you what the company knows or plans. If the competitor is a public company in the United States, you can read what it has told investors. Annual reports on Form 10-K include a section on risk factors and a discussion of the business by management. Companies describe competition, dependence on app store platforms, product risks and sometimes known technical problems there. The SEC's investor education site explains how to read a 10-K if you have not done it before.
You can search filing text for free with EDGAR full-text search on the SEC site, or start from the EDGAR company search page. Search the company name together with the word that describes your cluster, such as "sync", "outage" or "customer support". Three outcomes are possible.
The company mentions the problem. Then you know the gap is real and acknowledged. You can also read how they frame it and whether they describe plans to fix it.
The company is silent about it. That does not mean the gap is fake. Filings discuss material risks, not every bug. Silence just means the reviews are your only source.
The company describes the opposite. For example, it names that feature as a strength. That is the most interesting outcome. Either the users are wrong, or the company sees its product differently from the people who use it.
If the competitor is private, there is no 10-K. But a private company raising money from investors often files a Form D notice with the SEC. That filing does not describe the product. It does tell you the company is raising capital, and roughly when. A competitor with a clear weak spot and fresh funding may be about to fix it. We covered the fields in detail in how to read a Form D filing, so we will not repeat them here.
Step 8: turn the gap into a decision
A weak spot is only useful if it changes what you do. Write it down as one sentence: "Users of app B keep losing data when they switch devices, and that complaint has persisted across several releases." Then ask three questions.
Do our own users complain about the same thing? Run the same tags on your own reviews. If you share the weakness, it is not an advantage yet.
Can we make the difference visible? A gap only helps you if prospective users can see it. That can mean a clear feature, a comparison page, or a migration path for people who want to leave the other app.
How long will the gap last? A problem the competitor has acknowledged in a filing, backed by new capital, may close soon. A problem they describe as a strength may stay open for a long time.
What this method cannot do
Reviews come from the loudest users, not a random sample. Happy users write less. Some markets and age groups leave far fewer reviews than others. A complaint cluster tells you where friction lives. It does not tell you how many users feel it or how many left because of it. Keep your conclusions at that level. "This is a persistent, specific complaint that the other app does not attract" is a fair claim. "Most of their users hate this" is not.
The method also ages. Stores change how they sort and show reviews. Apps change fast. Rerun the comparison every few months rather than trusting one snapshot.
The short version
Pick two apps that do the same job. Tag the complaints by product area, not by mood. Compare shares, not counts. Read the full cluster for any gap. Discard obvious fakes. Then check the gap against what the company says in its own filings, or against signs that it is raising money to fix it.
If your competitor is a private startup, a funding round is often the first public sign that a known weak spot is about to get fixed. The Startup Capital Raises Form D report lists new Form D filings so you can see when a rival in your category has started raising money.
Want the signal instead of the raw filings? Get a free report preview. Prefer the tool to the write-up? Browse all data feeds or connect the free MCP server.