How to Set Up a WhatsApp Business Chatbot (2026)
Step-by-step guide on how to set up a whatsapp business chatbot — cover API access, bot builders, message templates, and what not to skip.
If you've been searching for how to set up a WhatsApp Business chatbot, you've probably already noticed the process is less straightforward than it looks. A WhatsApp Business chatbot can answer customer questions around the clock, qualify leads before your sales team gets involved, and send order updates automatically. But the setup has real gotchas — API approvals, message templates, opt-in rules — that trip up most first-timers. This guide walks you through every step, practically, with nothing glossed over.
What you're actually choosing between
Before you touch any settings, you need to understand that knowing how to set up a WhatsApp Business chatbot starts with picking the right entry point. "WhatsApp chatbot" actually refers to two different things, with very different setup paths.
WhatsApp Business App — the free app you download on Android or iPhone. You can set up quick replies, an away message, and a catalog. There's no true chatbot here, no API, and no way to connect an AI. It's fine for a one-person shop handling chats manually. It hits a wall the moment you want automation, multi-agent handling, or any kind of intelligent response.
WhatsApp Business API (now called Cloud API) — this is what enables real chatbots. It lets you connect WhatsApp to external software: bot builders, CRMs, helpdesks, or custom code. It requires a Meta Business account, phone number verification, and for most businesses, a third-party platform to sit between Meta and your actual bot logic. This is the path this guide covers.
Pick the app if you're testing the waters with a handful of chats per day. Pick the Cloud API if you want automation, AI responses, or scale — and if you're serious about how to set up a WhatsApp Business chatbot that actually handles real customer volume, the Cloud API is the only path worth taking.
Prerequisites before you start
Getting these in place first saves you from hitting dead ends mid-setup.
- A Meta Business account (business.facebook.com). Your personal Facebook account won't do.
- A phone number not already on WhatsApp. This is the number that will power your business chatbot. It can be a landline — WhatsApp will verify it via a call. Once used for the API, it can't be on the regular WhatsApp app simultaneously.
- A verified Meta Business Manager account. Meta will ask for your business name, address, and sometimes documents. Verification can take one to five business days, so start early.
- A platform or developer access. Unless you're writing raw API calls, you need a tool that handles the WhatsApp Cloud API connection for you — more on this below.
No shortcuts on these. Skipping verification is the most common reason businesses get stuck for days.
Step 1 — Set up Meta Business Manager
Go to business.facebook.com and create your business account if you don't have one. Fill in your legal business name exactly as it appears on your GST certificate or business registration — Meta cross-checks this during verification.
Under Business Settings → Security Center, start the business verification process. You'll need:
- Business phone number
- Business email on your own domain (not Gmail or Yahoo)
- Possibly legal documents depending on your country
India-based businesses: Meta accepts GST certificates, company incorporation documents, and utility bills in the business name. Allow 3-5 business days for review.
Once verified, your account gets elevated permission levels and you unlock higher messaging tiers (how many unique users you can message per day).
Step 2 — Create your WhatsApp Business App on Meta
Inside Meta for Developers (developers.facebook.com):
- Click "My Apps" → "Create App"
- Choose "Business" as the app type
- Give it a name (internal label — users won't see this)
- Link it to your Business Manager account
- From the app dashboard, find "WhatsApp" in the product list and click "Set up"
This drops you into the WhatsApp setup flow, where you'll add a phone number and get a temporary access token to start testing.
Adding and verifying your phone number
In the WhatsApp section of your app, go to "Phone numbers" → "Add phone number." Enter the number you set aside, then choose voice call or SMS for the verification code. Landlines must use voice call.
Once verified, you'll see it listed with a status of "Connected." You now have a working WhatsApp Business API endpoint.
Getting a permanent access token
The dashboard gives you a temporary token that expires in 24 hours — fine for initial testing, useless in production. For a permanent token, go to Business Settings → System Users, create a system user with "Employee" role, assign your app with "Full control," and generate a token. Store it securely; you'll need it in every API request.
Step 3 — Choose your chatbot platform
This is the decision that most affects how quickly you can build and how capable your bot ends up. When thinking through how to set up a WhatsApp Business chatbot, platform choice is often where people spend the most time deliberating — and rightfully so. You have three broad options:
| Option | Best for | Technical skill needed | Typical cost |
|---|---|---|---|
| No-code bot builder (e.g., respond.io, Wati, Interakt) | SMBs, quick launch | Low | $50–$200/month |
| Website chatbot with WhatsApp sync | Businesses already using a site bot | Low–medium | $9–$99/month |
| Custom code (direct Cloud API) | Developers, complex logic | High | Pay-per-conversation |
No-code platforms handle the authentication, webhook setup, and message template management for you. You just connect your WhatsApp number, drag flows together, and publish. The trade-off is cost and the fact that most of these platforms don't give you a knowledge-base-powered AI — they give you decision-tree flows, which break down the moment a user asks something outside the script.
If your goal is an AI-powered bot that actually understands your business content — not just a flow chart — you want a platform that combines a knowledge brain with messaging. That's where the website chatbot approach comes in.
Not sure which option suits your business? See how Alee compares to SiteGPT for a side-by-side breakdown of AI chatbot platforms for WhatsApp and website use.
Step 4 — Build your bot's knowledge base first
Here's something most guides on how to set up a WhatsApp Business chatbot skip entirely: the bot is only as useful as what it knows. A decision-tree bot with 20 pre-written branches will frustrate customers the moment they ask question 21. An AI bot trained on your actual content — your product pages, your FAQ, your return policy, your pricing — can handle the long tail of questions naturally.
Recommended approach: build the AI knowledge bot first, then surface it on WhatsApp (and your website simultaneously).
Alee does exactly this. You train it on your content sources — your website URL, PDFs, YouTube transcripts, FAQ text — and it creates a searchable knowledge brain. An LLM then answers customer questions grounded only in your content, with sources, so there's no hallucination. Once trained, the same bot can appear on your website widget and be connected to WhatsApp through an integration.
Start free at aleeup.com to set up your AI knowledge bot before wiring it to WhatsApp — the training step takes about 15 minutes.
The sources Alee can ingest:
- Website URL or sitemap — paste your URL and it crawls your content automatically
- PDFs and documents — product manuals, policy docs, pricing sheets
- YouTube transcripts — paste a transcript of your product demo or tutorial
- Manual text / FAQ blocks — for content not published anywhere yet
Once your knowledge base is ready, you have a bot that can answer thousands of variations of customer questions. Now you pipe that into WhatsApp.
Step 5 — Set up message templates (required for outbound)
This is where WhatsApp's rules differ sharply from other channels. On WhatsApp, you cannot send a customer an unsolicited first message unless they opt in AND you use a pre-approved message template. This is a deliberate spam prevention mechanism, and the rules are enforced hard.
Two conversation types:
- User-initiated (service conversations): The customer messages you first. For the next 24 hours after their last message, you can reply freely with any content. No template required.
- Business-initiated (utility and marketing conversations): You send the first message. Must use an approved template. Costs a per-conversation fee (set by Meta, varies by country).
Creating a message template
In Meta Business Manager → WhatsApp Manager → Message Templates:
- Click "Create Template"
- Choose a category: Marketing, Utility, or Authentication
- Name your template (lowercase, underscores, no spaces)
- Add your message body with variables like
{{1}}for dynamic values - Submit for review
Templates are typically approved within a few minutes for Utility and Authentication. Marketing templates can take up to 24 hours and may require additional justification.
Example utility template for order confirmation:
```
Hello {{1}}, your order #{{2}} has been confirmed.
Expected delivery: {{3}}.
Track your order here: {{4}}
```
Marketing templates require explicit opt-in from the customer before you can send them. No buying lists, no adding numbers scraped from the web.
Step 6 — Handle opt-ins properly
This is the step most businesses rush past, and it's the one that gets accounts flagged or banned.
WhatsApp requires documented, explicit opt-in before you send any marketing messages. The opt-in must:
- Clearly state the business name the customer is opting into
- Specify that they'll receive WhatsApp messages
- Be separate from other consents (can't bundle it into general terms)
Valid opt-in channels:
- A checkbox on your website checkout ("Send me order updates on WhatsApp — tick to opt in")
- A lead form where WhatsApp is clearly stated
- An in-person sign-up sheet
- A customer message to your WhatsApp number first (that's implicit opt-in for service replies)
For India-based businesses, TRAI regulations around marketing messages apply alongside WhatsApp's own rules.
Store your opt-in records — timestamp, channel, consent text shown — in case Meta audits. They do audit, and they suspend accounts that can't prove consent.
Step 7 — Wire up the webhook
Meta sends incoming messages to a URL you specify; your bot processes them and replies. If you're using a no-code platform, this is handled for you automatically.
If you're connecting your own server:
- In Meta for Developers → your app → WhatsApp → Configuration, find the Webhook section
- Set your Callback URL — the endpoint that will receive messages
- Set a Verify Token — a string Meta passes to confirm your endpoint
- Subscribe to the "messages" webhook field
Your server responds to the verification challenge Meta sends, and after that every incoming message arrives as a JSON payload you parse and act on.
A minimal webhook handler in Node.js looks like this:
```javascript
app.post('/webhook', (req, res) => {
const body = req.body;
if (body.object === 'whatsappbusinessaccount') {
body.entry.forEach(entry => {
const changes = entry.changes;
changes.forEach(change => {
const message = change.value.messages?.[0];
if (message) {
// send to your AI bot, get response, reply via Cloud API
handleIncoming(message);
}
});
});
res.sendStatus(200);
}
});
```
Step 8 — Connect your AI bot logic
This is where the WhatsApp pipe meets the intelligence layer.
When a message arrives, you pass the question to your knowledge bot, get the answer, and reply via the WhatsApp Cloud API. The experience is seamless — the user asks, gets a real answer in seconds.
With Alee's API, the flow is:
- Incoming message from customer arrives at your webhook
- You extract the message text and the sender's phone number
- POST the question to Alee's chat API along with a session ID (the phone number works as a session ID to maintain conversation context)
- Alee returns an answer grounded in your knowledge base
- POST the answer to WhatsApp Cloud API's messages endpoint with the sender's number as the recipient
Set it up once, and every customer question flows through automatically.
See how Alee works for a deeper look at the retrieval and answer generation under the hood.
Step 9 — Test before going live
Don't skip this. WhatsApp has quality ratings that affect your sending tier — too many blocks or reports early on can throttle your account badly.
Pre-launch testing checklist:
- Send test messages from your own phone number (added as a test number in the developer dashboard)
- Test every message template with all variable combinations
- Test edge cases: emojis, long questions, questions completely outside your knowledge base
- Confirm your bot sends a graceful fallback when it doesn't know the answer ("I don't have that info, but you can reach our team at...")
- Verify opt-in flows are working end-to-end
- Check that human handoff works when the bot escalates (if you have a live chat layer)
Run at least 30-40 test conversations covering common questions, edge cases, and the flows you expect in the first week.
Common mistakes that get accounts flagged or suspended
These come up constantly, and each one is avoidable. Many businesses learn about how to set up a WhatsApp Business chatbot only to get their account suspended in the first month because they rushed past these.
Sending to users without opt-in. Even if you're certain someone would want your message, no explicit opt-in means no message. One flagged contact can hurt your quality rating; many can trigger a review.
Using vague or misleading template names. Meta reviewers look at both the name and the body. A template called "promotion_update" that reads like spam will get rejected.
Not handling 24-hour windows properly. Many businesses reply outside the 24-hour user-initiated window with freeform text, which gets rejected by the API. Your bot logic must track when the window opens and closes and switch to templates accordingly.
Ignoring opt-outs. If a user sends "STOP" or blocks your number, remove them from all marketing lists immediately. No exceptions.
Running on personal WhatsApp simultaneously. Once a number is on the Business API, it can't be on the regular app. Trying to run both causes authentication conflicts and usually results in being locked out of one.
What a well-set-up WhatsApp chatbot actually handles
Once it's live, here's what a properly configured bot takes off your team's plate:
- Product or service questions — "Do you have this in size 12?" "What's the turnaround time for custom orders?" — the kind of stuff your team answers 40 times a day
- Order status updates — triggered automatically when order state changes in your system
- Lead qualification — ask a few key questions (budget, timeline, use case), score the lead, and either pass to sales or send a relevant resource
- FAQ handling — return policy, opening hours, payment methods, shipping zones
- Appointment booking — if integrated with a calendar tool via webhook
- Language switching — WhatsApp's user base spans dozens of languages; a well-trained knowledge bot can handle multi-language queries if your source content covers them
For more on specific use cases and how to structure your knowledge base for each, see tutorials.
Pricing reality check
WhatsApp Cloud API pricing has a few layers. As of 2026:
- Free conversations: Meta gives each business 1,000 free service conversations per month
- Beyond free tier: conversations are billed per 24-hour window, and the rate varies by country and conversation type (utility, marketing, authentication, service)
- Platform fees: if you use a no-code platform on top, add their subscription cost
- AI bot: if you're using an AI knowledge base platform, add that fee
For a small business in India handling 200-300 WhatsApp conversations a month, the total cost (Meta fees + a platform like Alee) can be under ₹2,000–3,000/month after the free tier. For a larger operation, model it out before committing.
Compare Alee pricing to understand what tier fits your conversation volume.
Key takeaways
- Knowing how to set up a WhatsApp Business chatbot properly starts with the Cloud API — the free WhatsApp Business app has no AI or automation capability.
- Meta Business Manager verification comes first; skipping it delays everything else.
- Message templates are mandatory for any business-initiated conversation. They must be approved before use.
- Opt-in consent is non-negotiable — document it, enforce it, and honor opt-outs instantly.
- A knowledge-base-powered AI bot handles far more question variety than a decision-tree flow. Build the knowledge brain before wiring the WhatsApp connection.
- Test with at least 30-40 scenarios before going live to protect your quality rating.
- The 24-hour window rule governs when you can reply freely vs. when you must use a template — your bot logic must respect this.
If you want the AI knowledge-base layer built in 15 minutes before you touch the WhatsApp API: sign up for Alee free, train it on your content, then use the integration to power your WhatsApp bot. The features page covers exactly what sources it can learn from and how the retrieval works.
---
Frequently asked questions
Can I set up a WhatsApp Business chatbot without coding?
Yes. No-code platforms like Wati, Interakt, and respond.io handle the Cloud API connection for you. You connect your verified phone number, build flows visually, and publish without writing code. If you also want an AI that answers from your knowledge base (not just pre-written flows), look for platforms that include a knowledge-base feature or integrate with one. For a purely knowledge-driven AI bot with a WhatsApp pipe, Alee trains on your content and outputs answers through an API you can connect to WhatsApp. With no-code tools, the full process can take an afternoon once Meta Business Manager is verified.
Do I need a WhatsApp Business API number that's different from my personal number?
Yes. The number you register with the Cloud API cannot simultaneously be active on the regular WhatsApp or WhatsApp Business app. Use a dedicated business number — a SIM you can receive an SMS or voice call on for verification, then port into the API. Many businesses buy a new SIM for this purpose.
How long does WhatsApp Business API approval take?
The phone number verification itself takes minutes. Meta Business Manager verification — required before you can go live at scale — takes 1-5 business days depending on your country and how quickly you provide the requested documents. For Indian businesses, GST certificates speed the process up significantly.
How much does a WhatsApp Business chatbot cost to run?
Two cost layers: Meta charges per conversation (a 24-hour window) — service conversations are often free up to 1,000/month, while utility and marketing conversations carry a per-conversation fee that varies by country. On top of that, you pay for your bot platform. A small operation might spend $20-50/month total. A larger business should model it out per country rate. See Alee pricing for the AI knowledge-base layer cost.
What happens if my chatbot doesn't know the answer?
A well-configured bot should have a graceful fallback — something like: "I don't have that information — let me connect you with our team." You can wire this escalation to a human agent queue in your helpdesk or provide a contact email. Always test edge-case paths before going live. For Alee, you configure a custom fallback message so the handoff feels on-brand. See more guides on fallback flow configuration.
Build your own AI chatbot with Alee
Train it on your site, embed it anywhere, capture leads 24/7. Free to start.