We checked 79 busy restaurants in Subang Jaya and PJ. 51 had not replied to any of their 10 newest Google reviews.

On 2026-10-20 we read the 10 newest Google reviews of 79 busy, open restaurants in Subang Jaya and Petaling Jaya. 51 of the 79 (65%) had not replied to a single one of them. These are not quiet places: the median restaurant had 879 Google reviews and a 4.4 rating.

01 / MethodWhat we measured

For each restaurant we read its live Google listing and its 10 newest reviews, using a paid service that reads Google Maps listings (Outscraper). For each review we kept three things only: the stars, the date, and whether the restaurant had replied. No reviewer names. No review text.

Across all 790 reviews, 144 (18%) had a reply from the restaurant.

02 / The stricter testThe fair objection, and the stricter number

"Maybe the reviews were new and the owner had not got to them yet."

For some, yes. For 9 of the 51, all 10 reviews were less than a week old. So we tightened the test. We counted only reviews that were more than 30 days old, so every owner had at least a month to reply. 55 restaurants had at least one review that old. 38 of those 55 had replied to none of them.

The number shrinks. The pattern does not.

03 / The patternWhat the split shows

Look at the table again. Restaurants did not spread out evenly between "replies sometimes" and "replies often". Most sat at one end or the other: 51 replied to none, 11 replied to nearly all.

It looks less like effort and more like a routine. Either somebody's job includes it, or it is nobody's job.

A Google review is a bit like a guest leaving a note on the table. Nobody has to answer it. But the next guest reads the note... and they read the silence too.

04 / LimitsWhat this does not say

• It does not say these restaurants ignore their guests. An owner may answer on WhatsApp, in person, or not at all on purpose. We only looked at Google.
• It does not say a reply brings more customers or a better ranking. We did not measure that, and we have not found anyone who measured it properly.
• 10 reviews is a small window. It stands in for the real question, "does this restaurant reply to Google reviews". A restaurant that replied to its 11th review counts as zero here.
• Whether a review had a reply comes from that same paid service, read once. We did not check each one by hand. A reply written after our read is not counted.
• We cannot say how FAST anyone replied. The data shows whether a reply exists, not when it was written.
• It is not "Malaysian restaurants". It is 79 busy restaurants in one part of the Klang Valley, picked from our own list. Chains with 5 or more outlets were left out.

05 / Our mistakeThe failure we own

We first wrote this up as "45 of 70". That was the number on our letter list, because 70 is how many letters we planned. But we had read 79. Quoting the smaller list would have been true and still wrong, because a reader assumes "we checked 70" means we checked 70. The rate barely moves (64% against 65%). The honesty does.