Integrations

How to Set Up WhatsApp Auto-Reply for Your Business: Qualify and Route Leads Automatically

By Ridhwan14 Oct 20268 min read

You can set up WhatsApp auto-reply for your business using the WhatsApp Business API and Make.com — automatically qualifying inbound leads, routing them to the right team member, and logging every conversation to your CRM, all within 90 seconds of the first message arriving. This guide covers the exact setup, step by step, for Singapore businesses.

Key Takeaways

  • WhatsApp auto-reply requires the WhatsApp Business API — the standard WhatsApp Business app does not support webhooks or automation beyond basic scheduled messages
  • Make.com connects your WhatsApp API number to any CRM, calendar, or Telegram channel in a single scenario — no code required
  • Adding a Claude or GPT-4o module extracts structured lead data (name, intent, budget, timeline) from unstructured conversational messages automatically
  • Every automated sequence must include PDPA opt-in consent before sending follow-up messages — this belongs in your first auto-reply
  • A complete auto-reply setup with AI qualification takes 3–5 days to configure and test for a Singapore SME

Why the WhatsApp Business App Is Not Enough

Most Singapore businesses start with the WhatsApp Business app — the free mobile app with basic auto-reply templates and labels. It handles low volume adequately. The problems start when enquiry volume grows:

  • Auto-replies are static templates, not personalised responses
  • There is no webhook — messages cannot trigger external workflows
  • Lead data cannot be pushed automatically to a CRM
  • You cannot run conditional routing (different responses based on what the prospect says)
  • There is no integration with booking systems, Google Calendar, or Telegram alerts

The moment you want to do something useful with an inbound WhatsApp message — qualify the lead, log it, route it, or book a call — you need the WhatsApp Business API. This is a cloud-hosted API from Meta that gives you full programmatic control over sending and receiving messages.


Step 1: Get WhatsApp Business API Access via Meta Developer Console

  1. Go to developers.facebook.com and create a Meta Developer account using your business Facebook profile.
  2. Create a new App and select Business as the app type.
  3. Add the WhatsApp product to your app.
  4. Under WhatsApp > Getting Started, register a phone number. For Singapore businesses, use a local +65 number — either a new SIM dedicated to WhatsApp Business, or migrate your existing WhatsApp Business number (note: migration erases chat history).
  5. Generate a System User Access Token from your Meta Business Manager under Business Settings > System Users. This is a permanent token (unlike user tokens, which expire every 60 days) — use this for your Make.com integration.
  6. Copy your Phone Number ID — you will need this to send messages from Make.com.

Cost: The WhatsApp Business API uses a per-conversation pricing model. In Singapore, the typical cost is USD 0.04–0.06 per business-initiated conversation (your outbound messages) and USD 0.00–0.02 per user-initiated conversation (inbound messages from customers). For most SMEs, this works out to SGD 30–150 per month depending on volume.


Step 2: Configure the Inbound Webhook in Make.com

  1. Log in to Make.com and create a new scenario.
  2. Add a Custom Webhook module as the trigger. Make.com generates a unique webhook URL.
  3. In your Meta Developer Console, go to WhatsApp > Configuration > Webhook. Paste your Make.com webhook URL and set a Verify Token (a string you choose, e.g. lumora-webhook-2026).
  4. Subscribe to the messages webhook field — this fires every time your number receives an inbound message.
  5. Back in Make.com, run the scenario once and send a test WhatsApp to your number. Make.com will capture the raw JSON payload — this shows you the structure of incoming messages (sender phone number, message body, timestamp).

The raw payload looks like:

entry[0].changes[0].value.messages[0].text.body → the message text
entry[0].changes[0].value.messages[0].from → sender's phone number
entry[0].changes[0].value.contacts[0].profile.name → sender's WhatsApp display name

Add a Router module after the webhook. This lets you branch the scenario based on message content — for example, routing "book appointment" messages differently from general enquiries.


Step 3: Add the AI Qualification Layer

This is where automation moves from basic auto-reply to intelligent lead qualification.

  1. In your Make.com scenario, add an HTTP module (for Claude via Anthropic API) or the OpenAI module (for GPT-4o) after the initial webhook parse.
  2. Write a system prompt that tells the AI to extract structured data from the incoming message:
You are a lead qualification assistant for a Singapore service business.
Extract the following from the customer message:
- Full name (if mentioned)
- Company or profession (if mentioned)
- Service interested in
- Budget range (if mentioned, else "not specified")
- Timeline (if mentioned, else "not specified")
- Lead quality: classify as HOT (ready to buy, clear intent), WARM (interested, gathering info), or COLD (too vague or irrelevant)

Return a valid JSON object with keys: name, company, service, budget, timeline, quality.
  1. Feed the incoming message text as the user input to the AI module.
  2. Parse the JSON output in Make.com using the JSON module — this gives you clean, structured variables you can use in downstream steps.

This single step transforms "Hi I want to know about your services" into a CRM-ready lead record with a quality score — without a human ever reading the message.


Step 4: Configure Routing Rules

Use a Router module with conditional filters to handle each lead type differently:

HOT leads (quality = "HOT"):

  • Send a Telegram message to your sales rep's personal Telegram, including the lead's name, service interest, phone number, and a WhatsApp deep link to open the conversation immediately
  • Create a new CRM row in Notion or HubSpot with status "Hot Lead — Action Required"
  • Send the prospect a WhatsApp message: "I've just flagged your enquiry to our team — someone will be in touch within 15 minutes."

WARM leads (quality = "WARM"):

  • Send the prospect a Calendly booking link: "We'd love to set up a quick 20-minute discovery call — here's our booking link: [link]"
  • Create a CRM row with status "Discovery Call Sent"
  • If they don't book within 48 hours, trigger a follow-up WhatsApp message (handled by a separate nurture scenario)

COLD or unqualified:

  • Send a generic response with a link to your website or resources page
  • Log to CRM with status "Low Priority"
  • No further automated outreach

This routing logic ensures your sales team only sees pre-qualified, high-intent leads — not every "Hi" that comes through.


Step 5: Add PDPA Consent to the First Message

Under Singapore's PDPA, you must obtain explicit consent before sending follow-up marketing or sales messages to a contact. Build this into your very first auto-reply — the one that fires within 90 seconds of the initial message.

Example first auto-reply:

Hi [Name]! Thanks for getting in touch with [Business Name]. We've received your enquiry and will respond personally within the next few hours.

Before we do — do you consent to receiving follow-up messages from us about our services? Reply YES to confirm. You can opt out at any time by replying STOP.

Add a Router step after this message that checks whether the prospect replied "YES" before continuing any further automated outreach. If they do not reply within 24 hours, send one reminder only — then stop.

Log the consent status (Consented / Not Responded / Opted Out) as a field on the CRM row. This creates an auditable consent record in case of a PDPA audit.


Step 6: Log Everything to Your CRM

Every inbound conversation — regardless of outcome — should create or update a record in your CRM. In Make.com:

  1. After the AI qualification step, add a Notion (or HubSpot, Airtable) module.
  2. Create a new database row with: contact name, WhatsApp number, message body, AI-extracted fields (service, budget, timeline), lead quality score, PDPA consent status, and the timestamp of first contact.
  3. Add the conversation source if known (you can pass a UTM parameter via your WhatsApp link to distinguish Google vs Facebook vs referral traffic).

A CRM that fills itself automatically — with no manual data entry — is the foundation of a scalable sales pipeline. Within three months of running inbound automation, you will have clean data on which channels drive your best leads, which services attract the most enquiries, and what qualification signals predict deal closure.


Connecting the Booking Flow

For businesses where the next step after qualification is an appointment (clinics, financial advisors, consultants), add a Calendly integration to the WARM lead path:

  1. Generate your Calendly embed link for the relevant appointment type.
  2. Send it in the WhatsApp message after the AI qualification step.
  3. Add a Calendly trigger module to a separate Make.com scenario that fires when a booking is confirmed — this creates the calendar event in Google Calendar, sends an email confirmation, updates the CRM row status to "Appointment Booked," and sends the Telegram alert to the assigned team member.

This closes the loop between WhatsApp inbound and your calendar — a prospect goes from sending their first message to having a confirmed appointment, without any manual intervention.


Common Mistakes to Avoid

Using a personal number: Running automation on a personal WhatsApp number violates Meta's Terms of Service and will result in a ban. Always use a dedicated business number registered through the official API.

Over-automating the conversation: The best inbound automation systems are designed to qualify and route — not to close deals. The moment a prospect says "Can I speak to someone?" or expresses frustration, a human must take over. Build an escalation trigger that pauses automation and fires a Telegram alert immediately.

Ignoring message timing: Avoid sending automated messages between 10pm and 8am Singapore time. Make.com has a built-in time filter you can add before the send module.

No fallback for unexpected messages: Not every inbound message fits your qualification flow. Build a catch-all router path for messages the AI cannot classify, and route these directly to a Telegram alert for manual handling.


Ready to Build This for Your Business?

Building a Make.com WhatsApp auto-reply scenario from scratch takes 3–5 days if you are familiar with the tools, or 2–3 weeks for a full production setup with AI qualification, routing, and CRM integration.

If you would prefer to skip the build and have it done for you, Lumora builds end-to-end WhatsApp inbound automation systems for Singapore businesses. Our AI agent development services include full setup, WhatsApp Business API registration, PDPA compliance review, and 30-day hypercare.

Apply for a free build at lumoraaiconsultancy.com/free-agent/apply.


Want to understand the bigger picture of WhatsApp inbound automation before diving into the setup? Read our guide: WhatsApp Inbound Automation for Singapore SMEs. For clinics and healthcare providers setting up WhatsApp auto-reply specifically, see our medical clinic automation page. For multi-lingual auto-reply handling English, Singlish, Arabic, or Mandarin in a single agent, read our multi-lingual WhatsApp AI agent guide.

WhatsApp