Travel App Reviews 2026: Support Complaints Up 15%, Service Praise Down 22% Across 50,497 Google Play Reviews
Travel apps rarely lose customers inside the app. They lose them after the booking, when something goes wrong and nobody fixes it. What 50,497 reviews of 13 travel apps say about refunds, unreachable support, scam accusations, template replies, and the one failure each type of app keeps repeating.
Data: 50,497 Google Play reviews · 13 travel apps · 1 Jan – 11 Oct 2026
Everything below comes from 50,497 public Google Play reviews of 13 travel apps that Intura collected from the Indonesian Play Store with an English-language filter, posted between 1 January and 11 October 2026. Every number was computed from that export; how we did it is at the bottom of the page.
50,497
Google Play reviews analysed
13 travel apps · Jan – Oct 2026
+14.6%
Customer-service complaints
Q1 → Q3 2026, share of all reviews
−22.4%
Praise for helpful service
Q1 → Q3 2026, share of all reviews
20.3% vs 8.2%
“Scam” in negative reviews
money mentioned vs not
Prologue
What does a bad travel app review actually describe?
Usually something that went wrong after the booking, not inside the app. Take one review posted on 12 February 2026. A traveler had paid in full for a room through OYO. At 2:00 PM the hotel demanded an extra early check-in charge and wouldn’t hand over the room until it was paid. The traveler waited more than an hour on customer support, got moved to a second property, and was asked for the same charge there. Exhausted, they paid it.
Paid in full. Charged again at check-in. Moved to a second hotel. Charged again there.
233 other users marked that review helpful. Only 12 of the 50,497 reviews in our dataset reached that many, and among the 15,166 negative ones it ranks 6th. Hundreds of strangers read it and recognised their own trip.
That is the pattern this report follows. Across 13 travel apps and nine months of 2026, the complaints that dominate are about what happens when money is stuck and nobody answers: refunds, unreachable support, and bookings the property won’t honor. And on the measures we tracked, that gap got wider through the year.
Chapter 1 · The trend
Is customer service at travel apps getting worse in 2026?
On what reviewers write, yes. Negative reviews complaining about customer service went from 6.06% of all reviews in January–March to 6.95% in July–September, a 14.6% rise. Over the same months, positive reviews praising helpful service fell from 2.35% to 1.83%, a 22.4% drop. Both moves are statistically significant (chi-square, p < 0.01).
The two signals come from different people. One is counted in angry reviews, the other in happy ones. When complaints about service rise while compliments about service fall, our read is that service quality across these apps is slipping, not just the mood of one crowd.
Q1 vs Q3 2026, share of all reviews (excl. Airbnb)
Customer-service complaints +14.6% p = 0.004
Praise for helpful service −22.4% p = 0.003
Negative (1–2★) reviews +8.1% p < 0.001
Q1 = 12,961 reviews, Q3 = 13,526. Airbnb is excluded because its reviews start on 21 February. Chi-square test on each pair.
The overall mood follows. Excluding Airbnb, 1–2★ reviews were 29.8% of the total in Q1 and 32.2% in Q3. Month by month the line is noisy, with a spring peak in April and May, but no month after March dropped back to January’s level.
Share of reviews rated 1–2★, by month (excl. Airbnb)
Axis starts at 20%.
*October covers 1–11 October only (1,542 reviews).
Chapter 2 · The complaints
What do travelers complain about most in travel apps?
Refunds and customer service, by a wide margin, not bugs. Of 15,166 negative reviews, 21.7% describe a refund or cancellation problem and 19.9% describe a customer-service failure. Bugs, crashes and a slow app come to 7.0%. Travel apps mostly fail in operations, not in code.
Share of negative (1–2★) reviews, by complaint type
A review can describe more than one problem, so shares don’t add to 100. Keyword categories; treat each share as a floor.
| Complaint | Reviews | 1★ share | Median words | Avg thumbs-up |
|---|---|---|---|---|
| Refund or cancellation problem | 3,287 | 96.2% | 52 | 3.14 |
| Customer service failure | 3,025 | 94.9% | 50 | 3.41 |
| Bugs, crashes, slow app | 1,059 | 84.8% | 30 | 2.92 |
| Hidden fees or price changes | 654 | 89.9% | 51 | 5.72 |
| Login, account or verification | 575 | 90.8% | 35 | 2.38 |
| Property not as listed or dirty | 319 | 92.8% | 67 | 1.76 |
| Booking not honored at the property | 308 | 98.1% | 66 | 3.81 |
| All negative reviews | 15,166 | 89.9% | 25 | 2.00 |
What happens when a hotel doesn’t honor a booking?
It produces the angriest reviews in the dataset. Bookings refused at the property, overbooked rooms and reservations the hotel had no record of appear in only 2.0% of negative reviews (308). Yet 98.1% of them give one star, the highest share of any category, and they run 66 words at the median against 25 for negative reviews overall. Reviewers describe arriving late at night, being turned away, and then trying to reach support from the lobby.
Are hidden fees a common complaint about booking apps?
Not the most common, but the one other users agree with most. Hidden fees and price changes appear in 4.3% of negative reviews. Those reviews average 5.72 thumbs-up, nearly three times the 2.00 average for negative reviews. Two of the most-upvoted complaints in the dataset ask for the same thing:
“How hard is it to show me the total cost up front?!”
“JUST. SHOW. TOTAL. PRICE. ON. MAP!”
The Booking.com review is the second most-upvoted negative review of the 15,166. Price surprises are the complaint readers most often recognise from their own bookings.
Chapter 3 · The escalation
Why do travelers call booking apps a scam?
Mostly because their money got stuck, not because they found fraud. When a negative review mentions money (a payment, a charge, a refund, a fee), it calls the platform a scam, fraud or theft 20.3% of the time. When it doesn’t, the figure is 8.2%. That is 2.5 times the rate, across 5,367 and 9,799 reviews.
Negative reviews using scam, fraud or theft language
Bad reviews rarely describe a single failure. Of the 7,066 negative reviews our rules could categorise, 1,871 (26.5%) describe two or more problems. The most common pairing is a refund problem plus a support failure, in 1,004 reviews. The sequence reviewers describe is the same across apps: a booking or payment goes wrong, the customer can’t get help, the money stays stuck, and the conversation becomes a refund fight.
Our read: most of these reviewers are not describing literal fraud. They are describing a transaction that failed and a company that didn’t fix it. For the platform, the reputational damage is the same, and it lands at the exact moment the customer decides whether to book again.
Chapter 4 · The replies
Do travel apps reply to bad reviews, and how fast?
Most booking apps reply to nearly every review, some within minutes. Trip.com and OYO have a median reply time of about five minutes. Agoda and Traveloka reply in under half an hour. Booking.com replies to 98.6% of its reviews, at a median of 5.6 hours.
| App | Reviews replied | Median reply time | Template replies* |
|---|---|---|---|
| Trip.com | 99.1% | 5 min | 61.0% |
| OYO | 98.5% | 6 min | 0.6% |
| Agoda | 97.8% | 24 min | 33.1% |
| Traveloka | 98.2% | 26 min | 0.0% |
| tiket.com | 60.9% | 22 min | 8.8% |
| Expedia | 95.0% | 54 min | 56.5% |
| Booking.com | 98.6% | 5.6 h | 93.8% |
| Skyscanner | 19.3% | 8.7 h | 43.0% |
| Klook | 94.4% | 14.4 h | 0.0% |
*Share of replies to 1–2★ reviews whose exact text the same app used 10 or more times. Skyscanner replies to 19.3% of all reviews but 92.4% of negative ones.
Does a fast reply fix anything?
Speed is not resolution, and many replies are templates. Booking.com answered 3,154 negative reviews with 184 distinct texts. One of them appears under 1,094 negative reviews, and 287 of those were complaints about customer service itself. The reply tells them to contact customer service.
“Hello, we are sorry to hear that your experience was not satisfactory. We are always here to help and our Customer Service team is available 24/7. Please get in touch with us…”
Personalised replies can carry the same problem. Traveloka addresses reviewers by name, but 38 of its replies to negative reviews say it was “unable to find any case matching your description in our system” and ask the customer to email support. They sit under refund disputes, failed bookings and login errors alike. The customer has already said support failed them; the reply sends them back to support.
Which travel apps don’t reply to reviews at all?
Four of the 13 apps leave almost every review unanswered. Among the apps that do reply, complaint type barely matters: they answer 98.1% of negative reviews, and every complaint category sits between 97.9% and 100%. The real divide is between apps, not topics.
Airbnb
5 of 9,800
reviews replied
Access by KAI
17 of 1,249
reviews replied
Tripadvisor
19 of 1,278
reviews replied
RedDoorz
0 of 97
reviews replied
That matters most for Airbnb. Login and identity-verification problems make up 10.5% of its negative reviews, nearly three times the average, and those reviewers are often locked out of the app. Airbnb replied publicly to 5 of its 9,800 reviews.
Chapter 5 · App by app
Which travel app has the most negative reviews in 2026?
In this dataset, Agoda: 73.6% of its 4,308 reviews are 1–2★, with an average rating of 2.03. tiket.com (65.2%) and Access by KAI (53.7%) follow. Tripadvisor has the lowest negative share at 12.2% and the highest average rating, 4.41. These shares describe the reviews we collected, not each app’s overall Play Store rating.
| App | Type | Reviews | Avg ★ | 1–2★ | Signature failure (share of its negative reviews) |
|---|---|---|---|---|---|
| Agoda | Booking platform | 4,308 | 2.03 | 73.6% | Refunds 34.1% · support 29.5% |
| tiket.com | Booking platform | 279 | 2.40 | 65.2% | Support 26.4% · refunds 23.6% |
| Access by KAI | Train ticketing | 1,249 | 2.73 | 53.7% | Bugs, crashes, slow app 30.6% (4× average) |
| Traveloka | Booking platform | 1,191 | 3.26 | 42.0% | Refunds 23.4% · support 22.2% |
| OYO | Budget hotel chain | 5,644 | 3.59 | 31.3% | Room not as listed 8.0% (4× avg) · extra charges 7.6% |
| Klook | Activities | 826 | 3.83 | 28.0% | Refunds 28.6% |
| Booking.com | Booking platform | 12,160 | 3.85 | 26.3% | Support 26.0% · refunds 21.1% |
| Expedia | Booking platform | 6,699 | 3.90 | 25.0% | Refunds 22.9% · support 17.7% |
| Trip.com | Booking platform | 4,638 | 3.95 | 24.1% | Refunds 20.7% · support 13.1% |
| Airbnb | Home rental | 9,800 | 4.07 | 21.0% | Login & identity verification 10.5% (3× avg) |
| Skyscanner | Flight search | 2,328 | 4.18 | 17.5% | Few service complaints · support 5.9% |
| Tripadvisor | Reviews & discovery | 1,278 | 4.41 | 12.2% | Refunds 1.9% · support 2.6% |
RedDoorz (97 reviews) is too small to profile. Averages across all negative reviews: refunds 21.7%, support 19.9%, bugs 7.0%, login/verification 3.8%, property not as listed 2.1%.
Why does each type of travel app fail differently?
Because the complaint follows the business model. Apps that take the customer’s money and hand the trip to a third party (hotels, airlines, hosts) collect refund and support complaints. Apps that only search, or sell a single product they run themselves, fail somewhere else.
Complaint mix by business model, share of negative reviews
Booking platforms (OTAs) · 9,850 negative reviews
Lodging (OYO, RedDoorz, Airbnb) · 3,850 negative reviews
Activities (Klook) · 231 negative reviews
Search & discovery (Skyscanner, Tripadvisor) · 564 negative reviews
Train ticketing (Access by KAI) · 671 negative reviews
Booking platforms such as Agoda, Booking.com and Traveloka carry the refund-and-support burden: 25.7% and 24.1% of their negative reviews. Access by KAI sells its own train tickets, so there is no third party to blame and no refund chain to break; 30.6% of its complaints are about the app itself, some written in Indonesian despite the English filter (“lemot”, “sering error”). OYO’s signature is the room: 8.0% of its negative reviews say the property didn’t match the listing or wasn’t clean, about four times the average. Skyscanner and Tripadvisor don’t hold the customer’s booking, and refund and support complaints make up about 5% of their negative reviews each.
What it means
What should travel companies do about bad app reviews?
The data points to five moves, ordered by how much trust each one protects. Each one can be tracked month by month in the same review data used here.
Measure resolution, not reply speed
Trip.com and OYO reply to reviews in a median of about five minutes. Booking.com reused one text under 1,094 negative reviews, 287 of them complaints about customer service, telling the reviewer to contact customer service. Track whether the problem was fixed and how long it took. A fast reply that sends the customer back to the queue they just complained about counts as a miss.
Put money problems at the front of the queue
Negative reviews that mention money call the platform a scam 20.3% of the time, against 8.2% for those that don’t. A stuck refund or a double charge is where a complaint becomes an accusation. Route these cases to people who can issue money back, with a deadline the customer can see.
Confirm bookings with the property before arrival
Bookings not honored at check-in are 2.0% of negative reviews, but 98.1% of them are one-star and they run 66 words at the median, more than twice the typical negative review. Our read: a confirmation check with the property 24–48 hours before check-in costs far less than a stranded guest and a public review.
Show the final price first
Hidden-fee and price-change complaints average 5.72 thumbs-up, nearly three times the 2.00 average for negative reviews. The second most-upvoted negative review in the dataset (505 thumbs-up) asks Booking.com to show the total up front. Readers agree with price complaints more than any other kind.
Fix the failure your business model creates
Booking platforms fail on refunds and support (25.7% and 24.1% of their negative reviews). The train app fails on software (30.6%). Home rental fails on identity checks. A train app needs engineering time; a booking platform needs a service operation that can unstick a payment.
Disclosure: Intura builds social-listening and sentiment-tracking tools, the category the month-by-month monitoring above falls into. Read the recommendations with that in mind.
What this report doesn’t cover
- • Google Play only, Indonesian storefront, English-language filter. Many reviewers are international travelers, so the results describe English-language app reviewers, not Indonesian travelers as a whole, and not iOS users.
- • Review counts reflect how the data was collected, not market share. Agoda’s 73.6% negative share, for example, describes the reviews in this export, not the app’s overall rating.
- • Complaint categories are keyword rules, not human labels. A hand check of 70 labels found 57 correct (81%); login and account labels were the weakest (6 of 10). The categories catch 46.6% of negative reviews, so treat each share as a floor, not a full count. The rest are mostly short reviews such as “worst app”.
- • The Q1 vs Q3 trends exclude Airbnb because its reviews only start on 21 February. They show change in what reviewers wrote, not change in the service itself.
- • Statements about any company describe what reviewers reported. We did not verify individual claims, and a review is one customer’s account.