Personalised Community Building at an Impersonal Scale

When you’re working with thousands of young people, it’s easy to start sounding like a broadcast channel. Messages go out. Some are read. A few people respond. Most don’t.

That’s where this journey began.

The program Impactyaan partnered with had a clear purpose: engage young people in civic action while helping them build essential 21st-century life skills, observation, reflection, collaboration and ownership. The intent was strong. The community was large. But engagement felt uneven and transactional. Young people were present in the system, but not always participating.

The real challenge wasn’t Scale. It was Connection.

A student who had never taken an action received the same nudge as someone who had already contributed multiple times. Data lived in different places, making it hard to answer basic questions. And every additional engagement push meant more manual effort.

The question became: How do you solve this?

Step one: Fix the foundation before adding intelligence

Before introducing anything “AI-powered,” we went back to basics. All critical data—user attributes, actions taken, engagement signals were brought into one common database.

Simple dashboards replaced guesswork.

Instead of asking “Are people engaging?”, we would now ask better questions:

• Which cohort is most active this week?

• Where are users getting stuck?

• What kind of actions lead to repeat participation?

This clarity changed how decisions were made—faster, calmer, and grounded in reality.

From “everyone” to someone like you

With clean data in place, the real shift began. Users were segmented not by static labels, but by behaviour. Users who had just joined started on onboarding journeys while users who had already taken action were nudged to go deeper or reflect. Even within the active community, the messaging became different basis levels of consistency of action taking

Same system. Different experience.

Introducing an AI bot that could actually listen

TheAI-powered conversational bot wasn’t just about automation. It was about empathy at scale.

Users could respond in text—or simply send a voice note in their own language. 

Within a month, 30% of users chose voice.

That number mattered—not because it was high, but because it signalled trust. People were choosing the most natural way to express themselves.

Scaling engagement without scaling the team

As participation grew, operations didn’t spiral.

A structured communication calendar, a reusable message library, and behaviour-triggered nudges meant the system worked with the team, not against them.

For instance, when a user shared an action, an automated message followed—acknowledging their effort and explaining what that unlocked next. No manual follow-ups. No missed moments.

So, What changed?

Within three months:

• WhatsApp delivery rates stayed strong at 91%, with 65% read rates

• The number of engaged users grew 4.5×

• Users taking desired civic actions increased 2.5×

But the biggest shift wasn’t just in numbers. The system stopped talking to young people and started speaking with them.

A note from Impactyaan

If this story resonated with you, it’s probably because you’re asking similar questions. At Impactyaan, we spend our time working alongside mission-driven teams to answer these questions—through data, design, and technology that serve people first.

If you’re building a community, running a program, or simply rethinking how engagement should feel, we’d love to connect. Even if it’s just a conversation to compare notes. Do write to us at contact@impactyaan.com.