In a country as large and diverse as India, policymaking often struggles to reach the very people it intends to serve. Despite the guidelines of the Pre-Legislative Consultation Policy (PLCP) in 2014 – which requires the government to publish draft policies for public comments – citizen participation has historically remained limited, inaccessible, and fragmented.
Civis, a Mumbai-based nonprofit, has been working since 2017 to change that. Their mission is clear: make public consultations accessible, understandable, and inclusive for every citizen. Over eight years, Civis has facilitated 1,000+ consultations, partnered with 40 government departments, and created a model of public participation rooted in trust, transparency, and technology.
But their transformation truly accelerated when they adopted Glific, an open-source WhatsApp chatbot builder for nonprofits. What began in 2021 as a small pilot with 2,500 respondents quickly grew into a nationwide consultation engine that today engages 420,000+ active WhatsApp citizens.
This case study draws from a detailed conversation between Deepak Nanda (Project Tech4Dev) and Atharva Joshi (Civis) to understand how Civis is building a more participatory public sphere, and how technology is enabling scale, inclusion, and trust.
Civis’ WhatsApp & Glific Impact
- 4,20,000+ citizens directly engaged via WhatsApp using Glific
- Growth from 2,500 users (2021) to 4.2 lakh+ today
- 96% of all consultation responses now come via WhatsApp
- 3.8 lakh+ responses on a single state-wide consultation (Vikasit Maharashtra)
- 35,000+ voice notes submitted, many over 2 minutes long
- Completion rate jumped to ~80% on WhatsApp (from ~50% on web)
- 3× higher engagement on WhatsApp compared to web platforms
What WhatsApp + Glific Enabled
- Participation without logins, downloads, or typing barriers
- Support for voice notes, images, PDFs, and vernacular text
- Targeted outreach using language, geography, and past responses
- Automated callbacks showing citizens how their feedback shaped policy
- AI-assisted analysis for sentiment, clause mapping, and issue clustering
- Two-bot architecture: one for live government consultations, one for AI pilots and experiments
Key Learnings
- WhatsApp is not just a channel—it is the primary gateway to participation
- Voice-first interaction dramatically increases inclusion
- Trust grows when feedback is acknowledged and closed-loop communication is built
- Open-source tools like Glific allow faster prototyping, flexibility, and scale
Key Challenges
- Onboarding users to share location and personal data
- Managing multilingual inputs at scale
- WhatsApp message limits and long approval cycles
- Rising costs and complexity for large AI-driven workflows
The Early Years: Building a Bridge Between Citizens and Policy
Public consultations were always meant to be democratic, but the reality was often far from it. Policies published online were dense, technical, and inaccessible. Only a handful of citizens ever commented. Civis stepped in as a translator and facilitator. The goal was not to chase large numbers, but to ensure inclusive and actionable input from citizens who will be directly affected by policies.
As Atharva explained in the interview, Civis’ role has always been to reduce friction between citizens and policy documents. “We try to bridge the gap between policy documents and people commenting on these policy documents,” he shared.
They launched a web platform, built a network of government partners, and attempted to scale consultations through digital channels. But even with strong foundations, participation plateaued. Logging in, downloading PDFs, and submitting feedback was still too laborious for most citizens.
“For the transgender rule back in 2022… almost 50-odd people commented… but it made such a great impact, wherein almost 5 million transgender people were able to access the benefits, which earlier many couldn’t access because the onboarding form required last name to be compulsory, and many transgender people do not associate with the last name.” – shares Atharva
Here, a seemingly minor form field—“last name (mandatory)”—was blocking lakhs of people from accessing benefits.
The team realized they needed something dramatically more accessible.
Why WhatsApp Changed the Trajectory
A breakthrough came when Civis began exploring WhatsApp as a channel. India’s deep mobile penetration and WhatsApp’s user-friendly design made it an obvious choice.
Atharva emphasized this shift clearly: “We realized people want to talk, and WhatsApp was a game changer… people already understand the UI”.
During the massive Vikasit Maharashtra public consultation, the team observed something striking:
- Web platform responses: 15,000
- WhatsApp bot responses: 380,000+
This simple but strategic decision made participation as easy as chatting, resulting in 96% of all responses being received through WhatsApp.
One of the most striking insights came from how people chose to participate. Nearly 1 in every 11 submissions was an audio message; over 35,000 citizens chose to speak rather than type, with each note averaging over two minutes.
This unexpected pattern demonstrated how comfort with everyday tools like WhatsApp can overcome barriers of literacy, typing skills, or digital hesitation. Citizens felt empowered to speak their thoughts in their own language, offering richer, more discerning, and authentic feedback.
The survey also revealed that citizens are not passive respondents; they actively engage when provided with accessible channels.
The conclusion was immediate: WhatsApp wasn’t just a tool -it was the gateway to participatory democracy.
Why Glific
Civis explored several vendors and finally chose Glific, an open-source platform built specifically for nonprofits.
Three reasons stood out:
1. Lightning-fast prototyping for government deadlines
Government cycles are tight; speed was essential.
Atharva recalled an early meeting:
“Within an hour we were able to sketch out how the survey tool would look and function”.
2. Flexibility after data collection
Glific allowed them to:
- Process voice notes
- Translate messages
- Write to Google Sheets
- Trigger webhooks
- Create AI-based flows
- Customize multilingual UX
This level of post-response computation was not possible through other tools.
3. Alignment with open-source values
Civis works a lot with open-source tools and encourages government adoption. Glific’s model aligned perfectly – technologically and philosophically.
Scaling Impact: From 2,500 to 420,000 WhatsApp Respondents
In 2021, Civis onboarded only 2,500 citizens on WhatsApp. Four years later, the number is 420,000+ WhatsApp respondents.
It today has 800,000+ total citizens engaged across all channels (WhatsApp hosting half of them), and on average 30 people respond to a policy document that is seeking public feedback. (This has been calculated from 1000+ public consultations.)
Civis actively reaches out to right stakeholders groups, from their 4L+ opted-in collection that are most impacted and empower their voices, towards inclusive and implementable policies.
And it intends to engage 1 million citizens by 2030.
The Metrics That Matter: Completion, Efficiency, and Quality
Civis meticulously tracks the performance difference between their web platform and WhatsApp.
1. Completion Rates
- Web: ~50%
- WhatsApp: ~80%
2. Engagement Volume
WhatsApp consistently sees 3x more engagement on any consultation compared to the platform.
3. Richer Input Formats
WhatsApp enabled:
- Voice notes
- PDFs
- Images
- Vernacular responses
This dramatically enriched the quality of feedback.
4. AI-Assisted Insights
Civis built pipelines to analyze:
- Sentiment
- Feedback type (grievance, suggestion, recommendation)
- Clause-wise mapping to the policy document
Building the Two-Bot Architecture
Civis now operates two carefully structured WhatsApp accounts:
Primary Bot – For Live Government Consultations
This is their main consultation bot where they engage citizens on laws and policies currently open for public feedback. All public-facing consultation flows run here.
Used to run all consultation flows with:
- Multiple policies running in parallel
- Time-bound 30-day cycles
- Mass outreach to targeted audiences
Secondary Bot – For Experiments and AI
This bot is used for government and partner deployments. They white-label it with their name/logo, share admin access with both teams, and host special projects or pilots without affecting the main consultation bot.
This bot is home to innovations such as:
- AI explainers for the annual Union Budget
- RAG-based policy Q&A
- Prototypes tested with small focus groups
- White-labeled deployments with partners
This architecture enabled Civis to experiment safely without disrupting official consultations.
Closing the Loop: A Human-Centered Communication Model, AI experiments and learnings
Participation isn’t just about collecting comments—it’s about showing citizens that their voice matters. Civis saw extremely strong repeat engagement once they introduced callback messages.
Atharva explained:
“We send personalized callbacks saying: this was your response, this is how it was translated into the policy”.
This “closed loop” model reinforced trust while encouraging continuous civic participation.
Beyond this, Civis has undertaken 10–15 AI pilots across Maharashtra, Punjab, and Odisha, alongside initiatives with key Central Government Departments.
They learned that:
- AI can expose missing sections in policies
- Governments are cautious about unanswered questions
- Costs can escalate at scale
- RAG systems must be carefully designed
However, AI opened up a fascinating new direction.
Civis is now building a policy evaluation tool that uses citizen questions to identify gaps before the policy is made public.
“We are trying to create a framework to evaluate how citizen-friendly a policy is… based on the questions people ask”.
Government departments have been surprisingly receptive to this idea.
Not everything worked smoothly — complex bot flows, low marketing conversions, WhatsApp’s 1,000-message limit, long government adoption cycles, and platform constraints all posed challenges, but each one led to meaningful product and program improvements.
“Tech becomes your core at the program-design stage. You can’t just replicate your program—you have to tweak it for what Glific can offer.”
The message is clear: technology is not an add-on; it’s a structural shift:
- Start small with one focused chatbot use case, then expand as you learn.
- Design for accessibility: prioritize voice notes, short content, and personalized callbacks.
- Treat WhatsApp as the primary entry point and use data pipelines + automation for scale.
- Personalize outreach using variables (language, past responses, interests) and maintain separate live + testing bot environments.
- Build trust by educating users beforehand and consistently closing the feedback loop.
Civis’ journey demonstrates what’s possible when policy, technology, and citizen voice come together. Powered by open-source platforms like Glific, Civis has transformed public consultations from a bureaucratic checkbox into a genuine democratic process—one where people can understand, participate, and influence change.
Leave a Reply