Setu is a Bengaluru-built payments app for India's small business owners — the kirana store, the tuition teacher, the two-person tailoring shop — who had been left out of the fintech industry's obsession with metro professionals. Bloom built the positioning, the Hindi-first content system, and the national launch campaign that took Setu from a founder's pitch deck to 9.4 million installs.
The situation we walked into.
India's fintech category had a metro problem. Every major payments app was designed, written, and marketed for the English-speaking professional in Mumbai or Bangalore. The country's tens of millions of small merchants — who did more daily transaction volume, collectively, than the entire white-collar segment — were an afterthought: a stripped-down 'lite' version of an app built for someone else.
Setu's founders had built the product correctly from day one — offline-first, voice-navigable, functional on the cheapest Android devices in circulation. What they had not built was a brand that told that story. The pitch deck used the same language as every competitor: 'seamless,' 'frictionless,' 'next-generation.' None of it meant anything to the audience it was supposedly for.
The brief had a hard deadline: a national campaign timed to a festival season when small-merchant spending and borrowing peak, with a single shot at the budget the founders had raised for it. There was no room for a campaign that needed a second attempt.
The work we did.
Positioning Around the Overlooked Majority
We repositioned Setu around a single argument: the small merchant is not the edge case of Indian fintech, they are the market. The brand stopped apologising for being built for kirana owners and tuition teachers and started saying so directly, in the first line of every piece of marketing.
The tone we built was respectful without being reverential — plainspoken Hindi that a shopkeeper would actually use, not the Hindi of a government form or a corporate translation. We tested every headline by reading it aloud in a real shop before it went into production.
Hindi-First Content System
Every piece of content was written in Hindi first and adapted into English and regional languages afterward — the reverse of how the category normally works. This changed what got made: idioms and examples that made sense in Hindi first, rather than English concepts translated into approximate Hindi.
We built a library of short, voice-first video explainers, since a meaningful share of the target audience was more comfortable with spoken Hindi than with reading dense app instructions in any language.
National Launch Campaign
We built the campaign around real merchants, not actors — filmed in their own shops, speaking in their own words about what the app changed for them. The media plan combined regional-language television, in-market activations at wholesale markets, and performance media timed precisely to the festival spending window.
We ran the campaign in eleven languages simultaneously, with a shared creative structure but market-specific casting, dialect, and cultural reference in each version.
Results, with context.
Installs attributed to the campaign within its first eight weeks, across all eleven language versions.
Hindi plus ten regional languages, each with market-specific casting and dialect, not a single translated master.
Growth in daily transacting merchants, comparing the 8 weeks post-campaign to the 8 weeks pre-campaign.
Share of total campaign installs originating outside India's eight largest metro markets.
"Investors kept telling us to talk like the fintech apps they knew. Bloom told us the opposite — talk like the merchants we actually serve, and let the category catch up to us. Nine months later, the category is trying to copy our language."
Our honest reflection.
The 'mass market' in India is not one audience speaking one compromise language. Eleven versions of one campaign outperformed a single Hindi-and-English master by a wide margin, because a shared creative idea still needs to arrive in a shopkeeper's own dialect to feel like it was made for them.
Writing in the target language first, rather than translating into it last, changes the actual content, not just the words. Ideas that work in Hindi are not always the ideas that work in English, and treating one as the master version of the other quietly discards the best material in whichever language comes second.