Hotels now answer online reviews in 3.1 days on average, down from 5.6 days in the third quarter of 2024, according to Shiji’s Q3 2026 Guest Experience Benchmark. Over the same period the share of guest mentions that are positive fell from 70.5% to 69.8%, so faster replies are arriving alongside slightly cooler guest sentiment.
Shiji published the benchmark on November 5 from its Reviewpro Reputation data, covering 11.9 million hotel reviews and 45.7 million guest mentions worldwide. Shiji sells Reviewpro Reputation, so the numbers come from a vendor that earns money when hotels take review data seriously. The release does not describe how the hotels in the sample were chosen, which is worth keeping in mind when comparing your own property against the averages.
What the Q3 benchmark reports
The Global Review Index, Shiji’s reputation score, rose to 87.0%, up 0.2 percentage points year over year. Beneath that headline, the proportion of positive guest mentions declined from 70.5% to 69.8%. Shiji’s own reading is that headline ratings alone are becoming an incomplete measure of the guest experience.
Response behaviour moved the most. The global review response rate reached 68.1%, which leaves nearly one in three reviews with no reply at all. Average response time across three-, four- and five-star hotels now sits between 2.9 and 3.2 days, against 4.7 to 5.4 days a year earlier. Shiji says the acceleration may reflect growing adoption of AI-assisted tools in reputation management, and it offers no data on which tools hotels are using.
Net Promoter Score rose from 55.74 to 56.26 globally. Five-star hotels recorded 62.29, while three-star properties declined to 50.89. Latin America posted the strongest regional gain, from 51.41 to 54.47. One figure carries its own warning from Shiji: five-star hotels in North America fell from 57.39 to 39.85, a result the company says “warrants further examination of sample size and property mix” before anyone draws conclusions from it.
Reading Q3 against Q2
Shiji’s Q2 2026 benchmark, released on July 16, gives a second data point, and Shiji does not set the two quarters side by side itself. Laid out from the two releases, the quarter-over-quarter picture reads as follows.
- Review index: 87.3% in Q2, 87.0% in Q3, a fall of 0.3 percentage points.
- Average response time: 3.7 days in Q2, 3.1 days in Q3.
- Response rate: 67.2% in Q2, 68.1% in Q3.
- Global NPS: 56.9 in Q2, 56.26 in Q3. The Q3 release calls the figure an increase because it compares with 55.74 a year earlier, but it sits below the Q2 reading.
- Asia review index: 90.2% in Q2, 90.1% in Q3, still the highest region.
The two releases compare against different base periods, and the hotel samples may differ between quarters, so the gaps are indications and not measured effects. They still show a consistent direction: replies got quicker twice in a row while the sentiment and satisfaction readings eased slightly or stayed flat. Faster replies have not yet shown up as better sentiment in Shiji’s data.
Where guests write their reviews
The second shift in the Q3 data is the channel mix. Ctrip doubled its share of global hotel reviews from 3% to 6%, and in Asia its share rose from 12% to 19%. In North America, Expedia’s share of reviews grew from 41% to 45%. Shiji’s Q2 release had put Booking.com and Google together at more than 62% of global review volume, and the Q3 release does not repeat that figure, so the Q3 split between those two platforms and the rest is not available from the release.
For a hotel, the practical meaning is that the platform where reputation is built depends on the market. An operator with Chinese inbound guests may now have a Ctrip review stream large enough to need its own reply workflow. A US operator with a strong Expedia mix is collecting a bigger share of its feedback on a platform whose reply tools differ from Google’s. The same dependence on platform-controlled guest data showed up when Booking.com stopped sending guest phone numbers to hotel systems, which left hotels with fewer direct routes to a guest once the booking was made.
Survey response is the weak link
The benchmark also separates asking for feedback from receiving it. Post-stay surveys reached a 57.1% email open rate and a 9.7% response rate, while survey-behaviour analysis recorded a 95.1% questionnaire completion rate. Once a guest starts a questionnaire, almost all of them finish. The loss happens before that point: fewer than one guest in ten responds. Hotels that fix survey length have little left to gain on completion, and the remaining lever is the invitation itself, meaning its timing, channel and sender. Native post-stay messaging tools such as the one Mews launched for hotels in October sit in that part of the stay.
What it means for operators
Independent hotels and small groups that reply to reviews by hand are the most exposed to the response-time shift. A 3.1-day average is now the figure a reviewer’s peers are meeting, and a property that answers in a week reads as slower than the market in front of every future guest.
Management companies and brand operators running dozens of properties face a different problem. Their global averages hide the spread by star rating and region that Shiji reports, from a 50.89 NPS at three-star hotels to 62.29 at five-star hotels. A portfolio dashboard that shows one score will not show where a segment is slipping, and the Q3 data says the segments are moving in different directions.
Revenue and marketing leads should note the channel change, because review volume at Ctrip and Expedia feeds the ranking and conversion on those same platforms.
What to ask your reputation vendor
Operators evaluating review and reputation platforms can use the benchmark as a checklist of questions for any vendor, Shiji included.
- Does the platform report share of positive mentions as well as the rating score, so a gap like Q3’s 0.2-point rise against a 0.7-point fall is visible?
- Can replies be tracked by platform, since Ctrip and Expedia are gaining share in the markets where they matter?
- What is the survey response rate on the invitation, and who controls the send timing?
- How are the benchmark peers selected, and can the property compare itself against hotels of the same star rating and region?
The practical test for any operator is the first number in this piece. If your average reply takes longer than 3.1 days, you are behind the global average. If it is shorter and your positive-mention share is still falling, speed has stopped being the thing to fix.

