Everyone says the same thing about indie launches: "Get on Twitter. Product Hunt will make or break you. Hacker News is where developers hang out."
All true. All saturated.
What nobody talks about: WhatsApp is the most frictionless distribution channel in existence.
Here's how I used it to bring 1000 people to a travel planning app without a single paid ad.
The Idea
I was building Sond — an app for groups planning trips together. The core insight: trip planning is synchronous. You and your friends need to decide on dates, destinations, budget, activities. It requires back-and-forth conversation.
One day, someone joked: "Why can't I just plan trips in WhatsApp?"
Why not? What if people could interact with Sond through WhatsApp?
- Send the bot a destination ("Paris")
- Get back flight options, hotel suggestions, itinerary ideas
- Share results with friends in WhatsApp
- Tap a link to open Sond for full planning
Zero friction. Everyone has WhatsApp. Everyone checks it constantly. No download required.
Building It (A Weekend Project)
I used Twilio's WhatsApp Business API. The setup:
- Twilio account + WhatsApp Business integration (~30 min)
- Simple Node.js backend (~2 hours)
- Prompt engineering for Claude (~2 hours)
The flow:
User WhatsApp → Twilio webhook → Node.js backend → Claude AI → response back to WhatsApp
The backend was simple:
const twilio = require('twilio');
const Anthropic = require('@anthropic-ai/sdk');
const client = new Anthropic();
app.post('/whatsapp', async (req, res) => {
const userMessage = req.body.Body;
const phoneNumber = req.body.From;
const message = await client.messages.create({
model: 'claude-3-sonnet-20240229',
max_tokens: 1024,
system: `You are a helpful travel planning assistant.
When users ask about destinations, suggest flights, hotels, and activities.
Always end with: "Plan the full trip in Sond: [link]"`,
messages: [{ role: 'user', content: userMessage }],
});
const twiml = new twilio.twiml.MessagingResponse();
twiml.message(message.content[0].text);
res.type('text/xml').send(twiml.toString());
});
The prompt was the real work — getting Claude to give useful travel suggestions while routing people to Sond.
Example interaction:
User: "I want to go somewhere warm and cheap in April"
Bot: "Great options:
🌴 Bali - Flights: $400-600, Hotels: $20-50/night, beaches and culture
☀️ Mexico City - Flights: $250-400, Hotels: $30-60/night, amazing food
🏖️ Cancun - Flights: $200-350, Hotels: $40-80/night, beach lovers
Plan the full trip in Sond: https://sond.app?utm_source=whatsapp"
The Launch
I posted in a few Discord communities:
"Built a WhatsApp bot for trip planning. Text it a destination and it gives you flight prices, hotel options, and itineraries. Try it: wa.me/..."
The wa.me/ URL is WhatsApp's share format — clicking it opens WhatsApp and starts a conversation.
I expected maybe 20 people to try it.
The first day brought 150 messages.
The Numbers (3 Weeks)
Week 1: 300 WhatsApp conversations, 45 Sond signups, 12 trips planned
Week 2: 520 conversations (word of mouth picked up), 240 signups
Week 3: 780 conversations, 350 signups, 110 trips planned
Total: 1000 conversations, ~400 new users, zero paid ads.
The cost: $0 in ads. ~$50/month in Twilio charges. ~5 hours of my time.
Compare that to Twitter (30-50 users if you're lucky), Product Hunt (100-150 signups on launch day if it lands), or Hacker News (months of credibility-building).
Why WhatsApp Works
Zero friction. Everyone has it. No signup, no download, no learning curve.
Constant attention. People check WhatsApp 50+ times per day. They see your response immediately. No algorithm deciding whether to show them.
Conversation feels personal. A WhatsApp message from a bot feels more intimate than a web page or Twitter mention. It's in your pocket. It's 1-on-1.
Built-in sharing. If the bot gives a great response, the user shares it with their group chat. No "click here to share" button needed.
Differentiation. In 2026, most indie builders still haven't tried WhatsApp. Twitter and Product Hunt are crowded. A WhatsApp bot is novel. People talk about it.
The Limitations
This only works for:
- Conversational products — things that benefit from back-and-forth chat
- Quick feedback loops — user needs a response in seconds
- Discovery products — helping users explore options (travel, food, shopping)
It doesn't work for:
- Creator/publishing platforms
- Productivity tools requiring complex UI
- Multi-step workflows needing a proper app interface
What I Learned
Existing platforms beat building. Everyone's on WhatsApp. Building a bot there beats building your own community from scratch.
Conversation is underrated. Most indie product pages have walls of text. A chat interface is more engaging.
Novelty matters. Even if the functionality is mediocre (mine was), the novelty of "oh cool, a trip planning bot in WhatsApp" got people to try it.
AI changed the game. Pre-Claude/GPT, a WhatsApp bot meant writing thousands of if/then rules. Now you prompt-engineer and it works.
For Your Next Launch
Don't default to Twitter and Product Hunt. Ask yourself:
- What platform is your audience already on, constantly checking?
- What's a novel way to distribute there that no one else is doing?
- Can you build a bot or tool that provides genuine value in 4 hours?
For SaaS, that might be:
- Slack bot for project management
- Discord bot for community engagement
- Telegram bot for notifications
- SMS bot for alerts
The best distribution channels aren't where the most people are. They're where your specific audience already spends their time, and where a new entry still feels novel.
WhatsApp was both. That's why 1000 people found Sond without me spending a dollar.