If you want to know what customers think of a company, Trustpilot, the app stores and YouTube comments give you three different crowds answering three different questions. Trustpilot collects reviews about a company as a whole, often after a purchase or a support contact, so it is strongest on service, delivery, billing and trust. App store reviews are written by people who installed a specific app and are reacting to a specific version, so they are strongest on product bugs, feature gaps and onboarding friction. YouTube comments are written by viewers of a video, not necessarily customers at all, so they are strongest on how the public perceives a brand, a launch or a creator's verdict, and weakest as a record of real usage.
The short rule: pick Trustpilot when the question is about the company, pick the app stores when the question is about the software, and pick YouTube when the question is about reputation and attention. If you own the business, the free official exports from Trustpilot's business dashboard, App Store Connect and the Google Play Console are usually the better choice for your own data. Scraping public pages makes sense when you need to look at other companies, compare many of them at once, or watch a source over time without an account on each one.
Who writes on Trustpilot
Trustpilot is a review site for companies. A reviewer picks a business profile and rates the experience from one to five stars, with a written explanation. There are two main ways a review gets there. The reviewer can go to the site on their own and write one, which Trustpilot calls an organic review. Or the business can send an invitation after a purchase or interaction, which produces an invited review. Trustpilot labels these differently on the page, and the labels matter when you read the data. Its own help material explains the distinction and how invitations work on the Trustpilot support site.
The reviewer population skews toward people who finished a transaction and had a clear reason to write. People who were delighted and people who were furious are overrepresented. People who were mildly satisfied are underrepresented, unless the company runs a steady invitation flow, which pulls the distribution closer to the actual customer base. That is why a profile with mostly invited reviews looks different from one with mostly organic reviews, even for the same company.
What this means for your questions: Trustpilot is good for anything that happens around the product rather than inside it. Shipping times, refund handling, customer support tone, subscription cancellations, hidden fees, account closures. It is weak for granular product feedback, because most reviewers describe an outcome, not a feature.
Who writes in the app stores
Apple's App Store and Google Play only let people review an app they have downloaded with the account they are signed in with. Apple documents how ratings and reviews work for developers on its ratings and reviews page. Google explains the developer side, including how to reply to reviews, in the Play Console help centre.
The crowd here is narrower than on Trustpilot. Every reviewer is a user of that app on that platform. Many reviews are triggered by an in-app prompt, which means the timing is chosen by the developer and often lands after a successful moment. Other reviews are written in anger after a crash, a forced update or a paywall. The text is usually short. The star rating carries a lot of the signal.
App store reviews carry metadata that Trustpilot reviews do not. On both stores you can usually see the app version the review was written against, and the date. That makes app store data the best source for version-level analysis: did the complaints about login spike after a certain release, did the rating recover after a fix, which features are mentioned most in one-star reviews. Developers can also reply, and the reply is visible, which tells you how the company handles complaints in public.
Where app store data is weak: it tells you nothing about the parts of the business that are not the app. A bank with a well-rated app can have a terrible branch experience. A retailer with a buggy app can have excellent delivery. If the question is about the company, the app store answers the wrong question.
Who writes on YouTube
YouTube comments are not reviews. They are reactions to a video, written by whoever watched it. Some commenters are customers. Many are not. Some are fans of the creator and are reacting to the creator, not to the product. Some are arguing with each other. The platform does not require any relationship with the company being discussed.
That makes YouTube useless as a measure of customer satisfaction and very useful as a measure of public perception. If a product review video goes up and the comments pile on about a price increase, you are watching sentiment form in real time. If an official brand channel posts a launch and the top comments ask about a feature that was removed, that is a demand signal you will not find on Trustpilot. Creator channels in a niche also surface the comparisons people actually make: which competitor gets named in the comments under a review of your product.
Volume on YouTube is uneven. A handful of videos carry most of the comments, and a video from a large creator can generate more text in a day than a company's Trustpilot profile gathers in a year. The quality of that text is lower. Expect jokes, off-topic threads and bots.
The honest limit: YouTube comments cannot be tied to a purchase, a version or a verified experience. Treat them as a listening post, not as evidence.
What a logged-out visitor can and cannot see
This matters for anyone collecting data without an account on each platform.
On Trustpilot, the public profile pages show the reviews, star ratings, dates, the invited or organic label, company replies and the overall score. You do not see the reviewer's email, the order reference used for an invitation, or any review that Trustpilot has removed or is holding for checks. Businesses on paid plans can see more in their own dashboard, including the list of invitations sent, but none of that is public. Trustpilot's guidelines for reviewers describe what is allowed in a review, which also tells you what gets removed and therefore never appears in a scrape.
On the Apple App Store web pages, the review list visible without signing in is limited and does not expose the full history. Inside the App Store app and in App Store Connect, developers see much more, including per-country breakdowns and the ability to reply. If you do not own the app, you are working from a partial public list and should say so in any analysis.
On Google Play, the web store page shows a sample of reviews with a filter for rating and device type, and you can load more by scrolling, but there is no guarantee you reach every review. The Play Console gives the developer the full set, with filters and a CSV export. Again, as an outsider you have a sample.
On YouTube, comments are public on any video where the owner has not turned them off. Replies are visible. Comment counts are shown. What you cannot see is anything on videos set to private or unlisted without the link, or comments the channel owner has held for review or deleted. Channels that disable comments give you nothing.
Free ways to get the same data, and when they are better
If the company or app is yours, use the official route first.
Trustpilot lets business accounts download their own reviews from the business dashboard. The format and the plan requirements change over time, so check inside your own account. For your own profile this is complete and includes data the public page does not show.
Apple gives developers their reviews in App Store Connect and through the App Store Connect API, which supports reading customer reviews for your own apps. Google gives developers a reviews export in the Play Console, with monthly CSV files in a cloud bucket, and a Reviews API for the last week of reviews. These are the right choice for your own apps. They are complete, free and include replies and version data.
YouTube offers the YouTube Data API, which can list comment threads on any public video. The commentThreads documentation explains the parameters. The API is free to use with a project in Google Cloud, within a daily quota. For a single video or a small channel the quota is more than enough, and the API is more reliable than any page scraper. The quota becomes a limit only when you want to track many channels or very large comment sections every day.
Manual export is also fine for small jobs. Copying a few dozen reviews into a spreadsheet by hand is faster than setting up any tool, and it forces you to actually read them.
Scraping the public pages earns its place in three situations. First, when the company or app is not yours and there is no dashboard to log into. Second, when you want to compare many companies or apps side by side, which no official export supports. Third, when you want a time series across sources, so you can see Trustpilot complaints, app store ratings and YouTube sentiment on the same timeline for a competitor or a market. In those cases you accept that the data is a public sample, and you document that.
Matching the source to the question
A few worked examples make the choice concrete.
If you are evaluating a supplier before signing a contract, read Trustpilot and look at the organic reviews separately from the invited ones. Pay attention to how the company replies to one-star reviews. Ignore the app stores unless the service is delivered through an app.
If you are a product manager tracking the effect of a release, use app store reviews filtered by version. Trustpilot will mostly be silent about your release, and YouTube will only notice if a creator covers it.
If you are in marketing and need to know how a competitor's launch landed, YouTube comments under the launch video and under the first reviews from creators will tell you more in a week than the other two sources tell you in a quarter. Then wait for the app store reviews to catch up for the real usage signal.
If you are doing due diligence on a market, you need all three on the same timeline, which is where scraped public data becomes the practical option.
Known gaps in every source
None of these sources is a census of customers. Trustpilot shows people who were asked or who were motivated enough to write. App stores show users who were prompted or provoked. YouTube shows viewers, who may never have bought anything. All three remove content, and you only ever see what survived moderation. All three change their page layouts and their rules, which breaks scrapers and shifts what is visible to logged-out visitors. Any analysis built on them should state the collection date, the source, whether the data came from an official export or a public page, and whether reviews with a particular label were included.
If you need Trustpilot reviews for companies you do not own, delivered as structured records with the star rating, date, review text, company reply and the invited or organic label, the Trustpilot reviews scraper on this site collects them from the public profile pages and charges per review returned, so a small comparison stays cheap and a large one scales without a subscription.
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.