Half of the biggest apps on Google Play reply to almost nothing
We measured 36 major apps across 10,800 public Google Play reviews. The median app answers 1.2% of them, and 15 answered none at all. The full table, the method, and every app we had to exclude are below, with no email required to see any of it.
Collected
Half the apps measured answer less than this.
42% of the sample had not a single reply.
When apps do answer, they answer fast.
Pooled across every review, skewed by a few outliers.
The pooled rate says 18.3% and the median app says 1.2%. Both numbers are correct, and the distance between them is the whole story. A handful of apps reply to nearly everything, which drags the average up, while most of the field replies to nothing whatsoever. There is no middle: only 6 of 36 apps cleared fifty percent, and 15 sat at exactly zero. Replying to reviews is not something companies do a little of. They either treat it as a channel or they ignore it completely.
Sorted by reply rate. Sample is the number of qualifying reviews examined per app, and 1-star is the reply rate on the reviews that most need one.
| App | Category | Sample | Reply rate | Median reply | 1-star rate |
|---|---|---|---|---|---|
| Booking.com | Travel | 300 | 97.7% | 7h | 95.5% |
| MyFitnessPal | Health | 300 | 96.7% | under 1h | 94.3% |
| Todoist | Productivity | 300 | 93.3% | 19h | 88.1% |
| Babbel | Education | 300 | 80.3% | 14h | 94.3% |
| Robinhood | Finance | 300 | 67.3% | under 1h | 64.9% |
| Dropbox | Productivity | 300 | 51.0% | 2.9d | 84.0% |
| Discord | Messaging | 300 | 43.0% | under 1h | 86.8% |
| Telegram | Messaging | 300 | 39.7% | under 1h | 39.4% |
| Yelp | Shopping | 300 | 26.7% | 22h | 66.4% |
| Canva | Productivity | 300 | 22.0% | 6.5d | 88.5% |
| Revolut | Finance | 300 | 14.7% | under 1h | 76.2% |
| Trello | Productivity | 300 | 12.0% | 6.7d | 20.5% |
| Tripadvisor | Travel | 300 | 3.7% | 7h | 34.8% |
| Headspace | Health | 300 | 3.3% | n/a | 3.4% |
| Bumble | Dating | 300 | 2.3% | n/a | 1.9% |
| Social | 300 | 1.7% | n/a | 2.8% | |
| Evernote | Productivity | 300 | 1.3% | n/a | 1.0% |
| eBay | Shopping | 300 | 1.3% | n/a | 3.0% |
| Slack | Productivity | 300 | 1.0% | n/a | 2.9% |
| Tinder | Dating | 300 | 1.0% | n/a | 2.2% |
| PayPal | Finance | 300 | 0.3% | n/a | 0.0% |
| SoundCloud | Music | 300 | 0.0% | n/a | 0.0% |
| Shazam | Music | 300 | 0.0% | n/a | 0.0% |
| Social | 300 | 0.0% | n/a | 0.0% | |
| Social | 300 | 0.0% | n/a | 0.0% | |
| Airbnb | Travel | 300 | 0.0% | n/a | 0.0% |
| Uber | Travel | 300 | 0.0% | n/a | 0.0% |
| Notion | Productivity | 300 | 0.0% | n/a | 0.0% |
| Strava | Health | 300 | 0.0% | n/a | 0.0% |
| Calm | Health | 300 | 0.0% | n/a | 0.0% |
| Nike Run Club | Health | 300 | 0.0% | n/a | 0.0% |
| Duolingo | Education | 300 | 0.0% | n/a | 0.0% |
| Amazon Shopping | Shopping | 300 | 0.0% | n/a | 0.0% |
| Etsy | Shopping | 300 | 0.0% | n/a | 0.0% |
| Candy Crush Saga | Games | 300 | 0.0% | n/a | 0.0% |
| Clash of Clans | Games | 300 | 0.0% | n/a | 0.0% |
Pooled across all 36 apps, so this is the behaviour of the field rather than of any one company.
Two figures dominate this topic, and ours does not replace either. Read them together.
AppFollow, June 2026 measured 51.5 million reviews across 22,800 apps and found a 24.8% average reply rate. That is a far larger sample than ours and you should treat it as the authority on the number it reports. Two things make it a different measurement: it covers mobile games only, and 24.8% is an average across all reviews rather than a typical app. Our equivalent all-review figure is 18.3%, which is not far off theirs once you account for the fact that we sampled across 11 categories rather than gaming.
Appbot, November 2016 found that 97% of Google Play reviews went unanswered, from 3.5 million reviews. That figure is still quoted constantly and it has never been publicly updated. It is now nine years old.
So the honest framing of this page: it is smaller than AppFollow's and newer than Appbot's, it covers categories outside gaming, and it is the only one of the three you can reproduce yourself, because the source data is public and the method is written out below. Where it disagrees with a bigger study, prefer the bigger study.
One number here has no counterpart in either: the median app replies to 1.2% of its reviews. An average tells you what happens to the typical review. A median tells you what the typical app does, and those turn out to be completely different stories.
Google Play publishes the developer's reply next to every public review, so an app's reply rate is observable by anyone. Nothing here is self-reported and none of these companies were asked. For each app we pulled reviews newest first, kept the 300 most recent that qualified, counted the ones carrying a reply, and divided.
A review qualifies if it has body text and is older than 7 days. Text matters because a bare star rating gives a developer nothing to answer, so counting those would turn the metric into a measure of how many silent raters an app has. The 7-day window matters because a review posted yesterday has not been ignored yet.
That window is measured rather than assumed. Across the 1,840 replies observed here, 59% arrived within twelve hours, 88% within three days, and 99% within seven. After a week the curve is flat, so a review still unanswered at that point is very unlikely to ever get an answer. Apps needed at least 100 qualifying reviews to appear at all, so no percentage here is computed off a handful of rows.
Reply latency is reported as a median, not a mean, so one reply arriving eighteen months late cannot move an app's figure.
Every app we tried and could not measure, with the reason. Silent truncation is how a benchmark starts lying, so these stay visible.
- Spotifyreview velocity too high to sample past the 7-day window (4000 reviews span 5.4 days)
- Instagramreview velocity too high to sample past the 7-day window (4000 reviews span 1.9 days)
- TikTokreview velocity too high to sample past the 7-day window (4000 reviews span 5.3 days)
- WhatsAppreview velocity too high to sample past the 7-day window (4000 reviews span 1.4 days)
- Wiseonly 0 reviews older than 7 days
Reply quality, for one. A single canned line pasted onto three hundred reviews scores identically to a considered answer, and reading the raw data, several of the high scorers are plainly templated. It also says nothing about who or what wrote the reply, and nothing at all about the App Store, where Apple's public feed carries no developer responses and this measurement is simply not possible without authenticated access to each app.
Character limits, edit and delete, permissions, publication delay, with the vendor quote next to every answer.
Read →Write inside Google Play's documented 350, with the App Store figure marked for what it is.
Read →Neither store claims it does. Here is what actually moves an average.
Read →Benchmark FAQ
The method, in the open.
How the numbers were collected, what the window is for, and what you should not read into them.
See how Reply Argus repliesHow was this measured?
Google Play publishes the developer's reply alongside every public review, so reply rate is an observable fact rather than something a company has to disclose. We sampled the 300 most recent qualifying reviews for each app, counted how many carried a reply, and divided. No Play Console access, no relationship with any of these companies, and nothing self-reported.
Why exclude reviews from the last week?
Because a review posted yesterday has not been ignored yet, and counting it would make every result depend on the hour we happened to run the collection. The 7-day window is not a guess: across the 1840 replies we observed, 88% arrived within three days and 99% within seven. Seven days is where the distribution flattens, so anything still unanswered after it is very unlikely to ever get an answer.
Why are Spotify, Instagram, TikTok, WhatsApp and Wise missing?
They receive reviews faster than a public sample can page backwards. WhatsApp's four thousand most recent reviews cover about a day and a half, so reaching past a seven-day window would mean pulling tens of thousands of records off an endpoint we do not pay for. Rather than quietly dropping them, every excluded app is listed below with its reason. An exclusion you can see is worth more than a number you cannot check.
Does this cover the App Store too?
No, and it cannot. Apple's public review feed carries no developer responses at all, so the same measurement is impossible without authenticated App Store Connect access to each app. Anyone publishing a cross-store reply-rate figure is either using private data or estimating. This is Google Play only, and the number should not be read as a statement about iOS.
Is this the biggest study of app review reply rates?
No, and it does not try to be. AppFollow published a study in June 2026 covering 51.5 million reviews across 22,800 apps, finding a 24.8% average reply rate. Theirs is a far larger sample and it is gaming only. Ours is smaller, spans categories outside gaming, and is reproducible from public data. The other figure people quote, Appbot's 97% unanswered, is from November 2016 and has never been publicly updated. Read all three together, and where ours disagrees with a bigger sample, prefer the bigger sample.
Why is your number so different from the 24.8% one?
Mostly because they are different statistics on different populations. 24.8% is an average across all reviews in a gaming-only sample. Our comparable all-review figure is 18.3%, which is in the same territory. The 1.2% on this page is a median across apps, which answers a different question: not what happens to the typical review, but what the typical app does. Those diverge sharply here because a handful of apps reply to almost everything and pull the average up while most of the field replies to nothing.
What does this not tell me?
Whether the replies are any good. A canned line pasted onto three hundred reviews scores the same here as a thoughtful answer, and several of the high scorers are clearly templated. It also says nothing about whether a human or an automation wrote them. Treat the rate as a floor on effort, not a measure of quality.
Can I use these numbers?
Yes, freely, including commercially, and there is no form to fill in. Cite the collection date, because these move. If you want to check the work rather than take it on faith, the method above is complete enough to reproduce independently, which is the only kind of benchmark worth quoting.
Most apps reply to nothing. Be the exception.
Reply Argus drafts an on-brand reply to every App Store and Google Play review, in the reviewer's own language, and waits for your approval before anything publishes.
Free plan · no card to begin