WhatsApp has become the default way Indian customers communicate with businesses. With over 500 million users in India alone, it is where your customers already are. Setting up a chatbot on WhatsApp sounds straightforward — but there is quite a bit involved before you get to a working, reliable system.
This guide covers everything you need to understand: the technical requirements, platform choices, realistic costs, and what can go wrong. The goal is to help you make an informed decision — whether you want to set it up yourself or work with someone who does this regularly.
What exactly is a WhatsApp chatbot?
A WhatsApp chatbot is an automated system that responds to your customers on WhatsApp without any human involvement. When a customer messages your business number, the bot reads the message, understands what they are asking, and replies — instantly, 24/7.
There are two types you will come across:
- Rule-based bots — follow preset decision trees. They work for simple, predictable questions like "What are your hours?" or "Where is my order?" but break down the moment someone asks something outside the script.
- AI-powered bots — use large language models to understand natural language. They can handle follow-up questions, context switching, and nuanced conversations. This is what most serious businesses are moving toward in 2026.
Most platforms marketed as "AI chatbots" in India are still rule-based underneath. They use keyword matching, not genuine language understanding. Always ask the vendor how the bot handles a question it has never seen before.
The WhatsApp Business API — the foundation everything depends on
You cannot build a chatbot on the regular WhatsApp Business app. For automation at any meaningful scale, you need access to the WhatsApp Business API, which is Meta's official platform for businesses.
Here is what that involves:
- Apply through a BSP (Business Solution Provider) — Meta does not give API access directly to most businesses. You go through an approved partner like WATI, AiSensy, Interakt, or Gupshup. They handle the API connection and provide a dashboard on top of it.
- Facebook Business Manager verification — your business needs to be verified on Meta's Business Manager. This requires a registered business, a Facebook page, and supporting documents. Can take 1-7 days.
- Phone number registration — the number you use for your chatbot must be registered with the API. It cannot be a number already active on regular WhatsApp or WhatsApp Business. Many businesses use a separate SIM for this.
- Message template approval — any message your bot sends first (outbound) must use a pre-approved template. Meta reviews and approves these templates, which typically takes 24-48 hours per template.
This setup process takes most businesses between 3 to 10 days depending on how quickly they gather documents and how responsive Meta's review process is. Rushing it is not possible.
Choosing the right platform
Once you have API access, you need a platform to build the actual bot logic. The main options used by Indian businesses in 2026:
| Platform | Best for | Starting price |
|---|---|---|
| WATI | SMBs wanting WhatsApp-focused features, team inbox | ~₹2,500/month |
| AiSensy | Marketing-heavy use cases, broadcasts | ~₹999/month |
| Interakt | Indian SMBs, Shopify/WooCommerce integration | ~₹999/month |
| Gupshup | Enterprise, custom development | Custom pricing |
| Custom build | Complex flows, deep integrations, AI conversations | One-time build cost |
The platform choice matters more than most people realise. A platform that is great for sending promotional broadcasts may not handle conversational AI well. One that works perfectly for customer support may not have the marketing automation features you need. Getting this wrong early means rebuilding later.
What does it actually cost in India?
WhatsApp chatbot costs have two components that most people do not realise at first.
Meta conversation charges
Meta charges per conversation (a 24-hour window), not per message. The rates for India as of 2026:
- Service conversations (customer initiates): free up to 1,000/month, then roughly ₹0.35-0.50 each
- Marketing conversations (you initiate): roughly ₹0.70-1.00 each
- Utility/authentication: roughly ₹0.15-0.30 each
Platform fees
On top of Meta's charges, your BSP platform charges a monthly subscription. This ranges from ₹999 to ₹25,000+ depending on the platform and plan.
A business handling 1,000 WhatsApp conversations per month should budget roughly ₹3,000 to ₹8,000 total per month — including platform fees and Meta charges. This is often significantly cheaper than the cost of having someone manually reply to messages.
What typically goes wrong
Most WhatsApp chatbot projects run into the same set of problems. Being aware of them upfront saves a lot of time and money.
- Using a personal number — the number you register cannot already be on WhatsApp. Many businesses discover this after starting the process.
- Template rejections — Meta rejects templates that are too promotional, too vague, or do not meet formatting guidelines. This delays go-live by days.
- Bot breaks on unexpected input — rule-based bots fail when customers ask anything outside the designed flow. Without proper fallback handling and human handoff, customers get stuck.
- No integration with existing systems — a chatbot that cannot check your actual order status or inventory is limited. Integration with your CRM, website, or ERP is usually where the real value comes from — and where complexity increases.
- Poor conversation design — technically working bots that frustrate customers because the conversation flow is unnatural or unhelpful. This is a design problem, not a technical one.
Should you set it up yourself or work with an agency?
This depends on what you are trying to build.
If you need a basic FAQ bot and broadcast system, platforms like AiSensy or Interakt are designed for non-technical users. With a few days of effort and their documentation, you can get something running.
If you need an AI-powered bot that understands natural conversations, integrates with your backend systems (order management, CRM, inventory), handles multiple scenarios intelligently, and has proper fallbacks to human agents — that requires either significant technical skill or a team that has built these systems before.
The honest answer is: the more you want the bot to actually do, the more complex the build becomes. A chatbot that just answers FAQs is a different product from one that qualifies leads, books appointments, checks order status, and escalates to your team at the right moment.
Want a WhatsApp chatbot built for you?
We build custom WhatsApp AI agents for Indian businesses — handling customer support, lead qualification, order tracking, and more. No generic templates, no one-size-fits-all.
Talk to us about your requirementsGetting started: what to decide first
Before you touch any platform or speak to any vendor, get clarity on these three things:
- What problem are you solving? Customer support backlog? Lead qualification? Order updates? The answer shapes everything else.
- What does the conversation need to do? Map out the 5-10 most common things customers message you about. That is your bot's job description.
- What systems does it need to connect to? If the bot needs to check real data — orders, appointments, inventory — it needs to integrate with your backend. Know this before you choose a platform.
With those three things clear, platform selection and build planning become much more straightforward.
WhatsApp automation in 2026 is genuinely powerful for Indian businesses — the engagement rates are far higher than email or SMS, and customers already prefer WhatsApp for business communication. Getting it right requires more thought than most people expect, but the results are worth it.