What scaling Setu from six crore to a hundred crore taught me about design’s job

In April 2025 I joined Setu, Pine Labs' fintech infrastructure company, as Head of Design. I still hold the role. In that time Setu's revenue went from around six crore to around a hundred crore. I was part of the leadership team that did that, and I want to be precise about what design's share of it was, because the honest version is more useful than the flattering one.

Design did not grow the revenue. Design grew adoption. Adoption grew the revenue.

What Setu sells, and why design matters to it

Setu sells infrastructure: Account Aggregator, identity verification services like eSign and KYC, and Bridge, the portal where merchants integrate and run it. Nobody buys infrastructure because it is beautiful. They adopt it when it is clear, trustworthy and quick to get working. Every one of those is a design outcome. A confusing consent screen is a lost Account Aggregator flow. An integration guide that hides the next step is a merchant who does not go live.

So the job was not to make Setu look better. It was to remove every reason a business might stop halfway.

Five people, then eight, and an operating system

I started with a team of five designers. The team is eight today. The number matters less than what sat around it, because a small team in a fast company gets pulled into whatever is loudest unless there is a system that decides where it goes.

Three mechanisms did most of the work.

Quarterly design planning. Every quarter, design sits with product leadership and lays out the upcoming initiatives, the company OKRs, and the design debt we are carrying. We map who takes what and when, and how each piece ladders to a company objective. It sounds like a calendar exercise. It is actually the moment design gets to say no with a reason: we are not taking that dashboard revamp this quarter because the Payments launch is the company's top priority, and here is the plan that shows why.

One intake for everything. Every request comes through one door, and the requester has to write down the problem and what success looks like before a designer is assigned. A surprising number of requests do not survive that question, and that is the point. The ones that do arrive with context, and we can look at fifteen requests against a capacity of ten and decide together which ten move.

A biweekly newsletter. What design shipped, what is in progress, what is coming, what we need. Two paragraphs, every two weeks, to everyone who works with us. It sounds trivial. It is the single thing that changed how the rest of the company treated design, from a queue you drop things into to a function that runs itself. One product manager told me it made him do his homework before engaging. That is the whole idea.

The AI initiative

Along the way, Setu decided AI needed to be a real part of the product rather than a slide. I drove that initiative: we built an AI vertical, put together an AI Champions team across functions, and ran a company-wide hackathon. The winning ideas went straight into the AI product roadmap. Design led it not because AI is a design topic, but because the question of what an AI feature should do for a merchant is a product question, and design was the function already asking it.

The brand

We also rebuilt Setu's brand identity in this period. I mention it last because it is the part people assume a design team spends its time on, and it was a fraction of the work. It mattered for one reason: a fintech that sells trust cannot look like it was assembled from defaults.

What I would tell another Head of Design

• Decide what design is for before you decide what it makes. At Setu it was for adoption. Everything followed from that.
• Build the operating system before you need it. Planning, intake, and a way of telling people what you did. A team of five with those runs like a team of ten.
• Measure the thing the business measures. I could tell you how many screens we shipped. The number that mattered was how many businesses went live.
• Take the org problems that are not yours. The AI initiative was not a design problem. It became design's because we were willing to hold it.

The revenue number is the company's. What design owns is that more businesses trusted the product enough to use it. That is the job, and it is enough.