In 2018, my co-founders and I made a bet that most people thought was too early.
We believed that the next 500 million Indian internet users would not behave like the first 100 million. They would not type. They would not read long-form content in English. They would not use products designed for urban, English-speaking, smartphone-native Indians.
They would speak. In their own language. And they would trust voices they recognised.
That belief became Bolkar — India's first voice-led Q&A platform. And building it taught me more about product than any framework, book, or MBA course ever could.
The Assumption Everyone Makes
When product teams talk about building for "the next billion users" — it often means making the existing product cheaper or translating the UI into Hindi.
That is not building for Bharat. That is localising a product built for someone else.
Real Bharat-first product thinking starts with a completely different question. Not "how do we adapt this?" but "what would someone build if they started entirely from the needs of this user?"
The Bolkar user was not a slightly-different version of the urban Indian app user. They were a fundamentally different person with fundamentally different needs, contexts, and behaviours.
They had a question — about health, farming, legal rights, government schemes, financial products — and no trusted way to get an answer. Google existed. But searching required typing in English, reading text, navigating links. The friction was enormous.
What We Learned About Trust
One of the most important things Bolkar taught me is that trust in a product is not the same across user segments.
Urban, tech-savvy users evaluate a product on functionality. Does it work? Is it fast? Is the UI clean?
Bharat users evaluate a product on trust signals that are completely different. Is the voice familiar? Does the person answering sound like they are from my community, my context? Do they understand the nuance of my situation?
This is why we invested so heavily in creator onboarding. Our 1,000+ active creators were not content machines. They were trust anchors. They were the reason a user in a small town in UP would trust the answer they received.
We applied gamification to retain our top creators — boosting day-30 retention from 60% to 74%. Not because retention frameworks told us to. But because we understood that if creators left, user trust collapsed. The product and the community were inseparable.
The Language Problem Nobody Talks About
Building a truly multilingual product — not just translated, but genuinely multilingual — is one of the hardest product challenges in India.
We built Bolkar across five Indian languages. Not by translating content, but by building separate creator communities in each language. The Hindi content felt like Hindi. The Tamil content felt like Tamil. Not because we ran text through a translation engine but because the creators were native speakers who understood cultural context.
This required rethinking almost everything. Content moderation. Search and discovery. Creator incentives. Quality signals. None of the global playbooks applied cleanly.
We co-developed AI tools for voice moderation with 85%+ accuracy in audio categorisation. Not because we wanted to build AI for its own sake. But because manual moderation at scale across five languages was not viable, and the existing tools did not understand our context.
The Metric That Mattered Most
Every product has a north star metric. For Bolkar, ours was not DAU or total audio content. It was day-7 retention.
Because day-7 retention told us whether a user had built a habit. Whether they had found enough value in their first week to come back without being prompted.
We built personalised notification systems that lifted day-7 retention by 2 percentage points. That sounds small. At a million users, it is not small.
The insight behind that system was deeply human. A Bolkar user who received a notification about a question in their language, from their region, related to something they had shown interest in — that was meaningful. A generic "Come back and see what's new" was worthless.
What I Would Tell Any PM Building for India
Stop thinking about Bharat as a scaled-down version of the urban Indian market.
Start thinking about the specific human in front of you — their language, their trust networks, their information needs, their constraints. Build from there, not from your existing product assumptions.
The best products for the next 500 million will not come from adapting global or urban-Indian products. They will come from people willing to start from scratch with a genuinely different mental model of who the user is.
We built Bolkar with this belief. It scaled to a million users. It got nominated by Google for Best App in India. Not because we were the smartest team, but because we were willing to stay genuinely curious about a user that the rest of the industry was only performing curiosity about.
Start with the human. Everything else follows.