Sri Lanka Hotels Reply to 1 in 10 Reviews - and Almost Only the Complaints (2026)

Every Sri Lanka guide tells you to read the reviews before you book. Almost none tell you what to do with the answer the reviews give you. So we went back to the same 291,887 verified guest reviews that power this site and asked a question nobody publishes: how often do Sri Lankan hotels actually reply to the people who write to them?

The answer is uncomfortable, and it is also one of the most useful things a traveller can know. A hotel that answers its complaints is a hotel that fixes them. A hotel that stays silent tells you, in advance, how it will treat a problem during your stay.

1. The headline: silence is the default.

Of 291,887 reviews with text, only 29,237 — 10.0% — carry any owner reply. Nine in ten guests who take the time to write something about their stay get no acknowledgement whatsoever. That is not the picture you get from scrolling a booking page, where replies are cherry-picked to sit under the reviews that make the property look good.

2. Hotels reply to complaints - and almost nothing else.

Break the corpus into score bands and the behaviour is unmistakable:

• Under 7/10 (a genuine complaint): 20.6% get a reply — 7,750 out of 37,706.
• 7–8/10 (fine, unremarkable): just 6.0% — 3,534 out of 58,783.
• 8.1–9.5 (good): 6.3% — 3,622 out of 57,414.
• 9.6–10 (perfect): 10.4% — 14,331 out of 137,984.

A roughly three-to-four-times gap separates the complaint band from the middle. The pattern reads like a fire-alarm reflex: reply fast when someone is angry, stay quiet when someone is merely satisfied. It is rational (a public complaint is reputational damage) but it leaves a strange hole where the ordinary guest lives. The vast middle of the distribution — the 7s, the 8s, the quietly pleased — is the most ignored group in the entire dataset.

There is a second, sharper version of this. Of all 29,237 replies in the corpus, 26.5% are replies to a complaint — disproportionate, since complaints are only 12.9% of all reviews. Hotels allocate a quarter of their total reply effort to the 13% of reviews that are negative.

3. The number that should bother hotels most: 79% of complaints go unanswered.

Complaints get more attention, yes — but not nearly enough. Of the 37,706 reviews scoring under 7/10, 29,956 — 79.4% — never receive a reply. So even in the one band where hotels bother to respond, four in five unhappy guests are met with silence. The reputational tool hotels reach for fastest is still left unused on the overwhelming majority of the reviews where it matters.

4. When hotels do reply to a complaint, they write a real answer.

The length data tells its own story. Replies to complaints average 308 characters — a proper paragraph, often name-checking the issue and offering a remedy. Replies to good (8.1–9.5) and middling (7–8) reviews average just 28 and 32 characters: a "thank you", a heart, a canned line. Replies to perfect 9.6–10 reviews land at 75 characters — still short, and clearly templated.

In other words, the effort is real when there is a problem, and theatre when there is not. If you are deciding between two properties with similar scores, look at whether complaints buried in their reviews have owner replies — and read how specific those replies are.

5. What guests actually complain about (and why a reply matters).

We scanned the 23,725 English-language complaints (reviews under 7/10 with written text) for the themes guests raise most:

• Room & facilities — 29.9% (rooms, bathrooms, hot water, air-conditioning, Wi-Fi, pool, beds): the single biggest theme, and the one most likely to be fixable in a day.
• Staff & service — 26.5% (front desk, manager, host): the theme where a single owner reply can change how the whole review reads to the next guest.
• Breakfast & food — 18.1%, matching our 353,633-review breakfast study showing food is the most-discussed hotel feature overall.
• Location & transport — 16.6% (distance, tuk-tuk access): the one theme owners genuinely cannot fix — and the one where a clear reply heads off the next disappointed booking.
• Cleanliness — 11.7%, value & price — 10.4%, smell, damp & pests — 8.6%, noise — 3.8%.

Notice how much of that list is operational and fixable: hot water, cleanliness, breakfast consistency, noise. A property that replies to complaints is signalling it reads them. One that never replies is telling you the review page is a shop window, not a feedback loop.

6. How to use this when you book.

• Don't just read the score - scan the low reviews and check for replies. A 9.0 with answered complaints is a safer bet than a 9.2 with a wall of silence.
• Read the reply length, not the reply counts. A 10-character "Thank you!" is decoration; a 300-character, specific response is management actually engaging.
• Weight the complaint themes against your own priorities. If you cannot do stairs, a "room & facilities" complaint is a red flag; if you'll be out all day, a "location" complaint may not concern you.
• Treat silence on complaints as a preview. The property that ignores a written complaint online is not more likely to solve your problem in person.

The bigger picture is in our full 291,887-review analysis, alongside the breakfast study, the Wi-Fi & working study, the nationality breakdown, and the destinations ranked by guest satisfaction. When you are ready to book on the numbers, start from our data-ranked list of Sri Lanka's highest-rated stays.

Methodology: 291,887 verified Booking.com guest reviews of Sri Lankan properties (10-point scale) from the BYC.lk corpus. "Reply rate" = share of reviews where an owner response is present and non-empty. Score bands: under 7 (37,706 reviews), 7–8 (58,783), 8.1–9.5 (57,414), 9.6–10 (137,984). Complaint-theme frequencies computed over the 23,725 English-language reviews scoring under 7/10 with written text; a review may mention more than one theme, so percentages do not sum to 100. Reply-length figures are mean character counts of owner responses within each band. Booking intent is influenced by many factors; this study measures published owner behaviour only.