How Zomato is creating the illusion of personalisation — without actually knowing you

There’s a very specific moment — usually somewhere between 1:10 PM and the existential dread of another workday — when Zomato pops up with “Your usual?” And for a second, you think: impressive, this app has a better grip on my life than I do.

It doesn’t. What it has is context. And context, when perfectly timed, can masquerade as intelligence.

Zomato didn’t build a psychological twin of you in their backend. There is no proprietary “Gyanesh cravings model v3.4.”

There’s just a neat combination of timing, recency, and a few behavioural tricks that make the whole thing feel far smarter than it is.

The Soft Science of “We Know You”

Zomato’s recommendations are not the product of deep personalisation — they’re the product equivalent of saying something generic with great confidence.

Show breakfast items in the morning. Surface biryani when the country is statistically hungriest. Label something “light dinner” and let the user pretend they make responsible decisions after 7 PM (LMAO).

The brilliance isn’t in precision. It’s in plausibility.

You don’t need to predict a specific user’s behaviour if you can predict the category’s behaviour and package it as individual insight.

Why This Works (And Why It Works Alarmingly Well)

Humans rarely question relevance when it lands exactly when they need it.

A reminder shown at the right time feels personal. A repeat of your last order feels like memory. Two suggestions feel curated, even if they’re just… two suggestions.

This is where PM intuition quietly outperforms machine learning:

people interpret timing as intent.

We are wired to assume that if something appears at the perfect moment, it must have been meant for us. That assumption does 80% of the heavy lifting.

The India Advantage

If you tried this in a country with limited cuisine variety and predictable eating habits, it might still work. But in India? It’s like running a con in a room full of willing participants.

Users aren’t loyal. They’re busy. And the app that reduces thinking wins.

Zomato’s not “personalising.”

It’s removing (or atleast trying to) friction in a market where attention is a scarce resource.

The Business Logic (A Very Sensible One)

These lightweight nudges yield heavyweight outcomes:

It’s operational efficiency wearing the coat of personalisation — a coat that costs significantly less than GPUs and research teams.

What PMs Should Actually Steal

The lesson isn’t “fake personalisation.” It’s this:

You don’t need deep data. You need useful data. Recency is cheap and effective. Time-of-day is a signal, not a suggestion. Reducing choice increases velocity. And perceived relevance is often enough to drive behaviour change.

If you can catch users at the exact moment they’re about to make a decision, you don’t need to “know” them. You just need to appear as if you do.

Zomato understands this. And the result is a product experience that feels personal without ever needing to be personal.

Another observation..

Have any of you used the sort feature in the Zomato app? It's buried deep down, while I'd expect it to be an essential part of Discovery. I'd love to hear if you share similar thoughts!