Not everyone like's to read to much nowadays (thanks TikTok). See a comic strip?

Not everyone like's to read to much nowadays (thanks TikTok). See a comic strip?

Not everyone like's to read to much nowadays (thanks TikTok). See a comic strip?

Comic Strip

Comic Strip

Comic Strip

Paytm Chat

Paytm Chat

Paytm Chat

Social Payments for 500M+ Monthly Users

Social Payments for 500M+ Monthly Users

2020 - 2021

Product Designer

Tools: Figma ·Storybooks

Tools: Figma ·Storybooks

$500M+

$500M+

Monthly Users

Touch this Homepage

Monthly Users

Touch this Homepage

6 months

6 months

Start to

Launch

Start to

Launch

The Problem

The Problem

The Problem

I joined Paytm in 2020, right as UPI; India's real-time payments rail was reshaping the market. Paytm had been the default way to send money in India for years. Then Google Pay integrated UPI and added something Paytm didn't have: you could chat with someone while paying them. Users started migrating.


Paytm's answer was to bring that same idea to its own homepage: a surface that over 500 million people touch every month. Whatever shipped there wasn't a feature update. It was a brand statement.

The Caveat

The Caveat

The Caveat

Every component we built had to go back into Paytm's core design system; which meant every one of them needed legal, business, and management sign-off before it shipped.

Every component we built had to go back into Paytm's core design system; which meant every one of them needed legal, business, and management sign-off before it shipped.

Every component we built had to go back into Paytm's core design system; which meant every one of them needed legal, business, and management sign-off before it shipped.

Paytm had never had chat before. There was no bubble style, no message-state pattern, nothing to extend: we were defining primitives for the first time, in a domain (payments) that's about as regulated as consumer software gets in India. That's a real constraint most chat products never face: a "send" button isn't just a UX decision, it's a compliance conversation.


So we built deliberately, in public, with each new piece: chat bubbles, image sends, video sends, money sends, combinations of all of them checked at every stage rather than designed in a vacuum and approved at the end. Message states got the same treatment: sent, delivered, read, and reacted-to, each with distinct, tested visual treatment.

The result

The result

The result

$500M+

$500M+

Monthly Users

Touch this Homepage

Monthly Users

Touch this Homepage

6 months

6 months

Start to

Launch

Start to

Launch

Post-launch, the team ran A/B tests through CleverTap with the research team to refine the experience further — I'd moved to Paytm's design systems team by that point, so I can speak to what shipped, not to how those later tests changed it.

Post-launch, the team ran A/B tests through CleverTap with the research team to refine the experience further — I'd moved to Paytm's design systems team by that point, so I can speak to what shipped, not to how those later tests changed it.

NOTE FROM FUTURE SAMARTH

NOTE FROM FUTURE SAMARTH

We designed every message state (sent, delivered, read, reacted) in real depth, and I still think that level of care is why the experience felt trustworthy fast in a category (payments) where trust is the entire game.


If I did it again, I'd push to stay through the A/B testing phase. I built the foundation and handed off before seeing what the data actually said about it, and that's the one part of this project I never got to close the loop on.

Always up for a conversation about UX, design systems or AI-native tooling

Always up for a conversation about UX, design systems or AI-native tooling

Always up for a conversation about UX, design systems or AI-native tooling

© 2026 Samarth Nayyar