A LocalCircles national survey found that 88% of Indians check ratings before choosing a business, and that 46% do not trust what those ratings tell them. The instinct is to treat the second number as cancelling the first. It does not. People are still checking. They have simply stopped taking the number at face value.
That is a specific behaviour, and it has specific consequences for what your listing needs to contain.
Why the distrust is rational
It was earned. Paid five-star reviews are cheap and widely sold. Ratings get bundled into delivery platform promotions in ways customers can see. Competitor review-bombing is common enough that most people have watched it happen to somewhere they like. A reader who discounts a 4.9 is not being cynical. They are applying a correction that experience taught them.
So the reader arrives already assuming some of what they see is manufactured, and starts looking for which parts are not.
What a sceptical reader looks for
Someone who has stopped trusting the average reads differently. They skip the score and go hunting for evidence that a real business is on the other end.
- Negative reviews that are still there. A page with no criticism at all reads as curated. One or two honest complaints make the rest credible.
- Replies that engage with specifics. "We are sorry for your experience" fifty times over proves nobody read anything. A reply that names the actual problem proves someone did.
- Reviews that describe something. "Good" is worthless. "The Gurugram branch opens at 9 but the counter is not staffed until 9:30" is worth more than the score.
- A steady trickle over time. Clusters look purchased. A few every week looks like a business that is simply open.
- Agreement across platforms. A 4.7 on one app and a 3.1 on another means at least one of them is wrong, and the reader will not guess in your favour.
Every item on that list is something you influence by how you operate, and none of it is something you can buy.
The cross-platform problem is worse here
An Indian consumer comparing restaurants may see a Google rating, a Zomato rating and a Swiggy rating for the same kitchen, with different averages on each because each platform counts different things. Delivery ratings absorb complaints about riders and packaging that have nothing to do with the food. Dine-in ratings do not.
A reader who notices the spread has no way to work out which one reflects the business, so they assume the lowest is closest to the truth. Reconciling those scores, or at least being able to explain the spread when someone asks, is a larger part of reputation work in India than in markets where one platform dominates.
What actually closes the gap
Nothing on this list is clever. That is rather the point, because the reader is specifically looking for effort that could not have been faked.
Reply to the bad ones, visibly and specifically. This is the single highest-value action available, because it is the thing the sceptical reader is scanning for and the thing the least effort goes into.
Leave the criticism up. Deleting what you can and reporting the rest produces a page that nobody believes. A handful of unflattering reviews with good replies underneath does more for you than a clean sheet.
Ask everyone, not the happy ones. Selective soliciting produces exactly the uniform positivity that triggers the discount in the first place.
Keep the gap between platforms explainable. You will not make the numbers identical. You should know why they differ and be able to say so.
The 46% are not unreachable. They are doing more work than the average reader before they decide, which means there is more surface for an honest business to be convincing on. A business that only ever optimised its average has nothing to show them.