Auto-replies and smart WhatsApp flows for sales and support

Greenbubble's flow builder turns WhatsApp into a real customer-engagement channel — auto-replies that know who they're talking to, sales flows that qualify and route, and support flows that resolve or escalate.

Three flow patterns that pay for themselves in week one

1. First-touch welcome + qualification

When a new contact messages you, send a branded welcome, then ask a single qualifying question ("are you looking for product info, support, or a quote?"). Branch the conversation: sales-qualified contacts get a calendar link to book a call; support contacts go straight to a human.

The power of this pattern is that it replaces an open-ended "How can we help?" with a decision the customer can make in one tap. The answer becomes a routing signal, and it is captured on the contact record for later.

2. After-hours auto-reply + ticket creation

Outside business hours, the bot replies with an SLA ("we'll respond by 9am"), captures the issue in 3 taps (category + urgency + summary), creates a ticket in your helpdesk, and queues it for the morning shift.

This does two jobs at once: it sets expectations so the customer does not feel ignored, and it converts a late-night message into structured data your team can act on at 9am instead of reading a wall of chat.

3. Post-purchase NPS + review request

Seven days after delivery, the bot sends a one-tap NPS survey, branches by score (9–10 → Google review link, 7–8 → support follow-up, <7 → ticket creation), and updates the contact record in your CRM.

The timing matters. Seven days is long enough for the product to arrive and be used, and short enough that the experience is still fresh. Branching by score keeps unhappy customers in a private support lane instead of on a public review page.

Built-in building blocks

• Triggers — first message, keyword match, contact tag, webhook event, schedule.
• Actions — send template, send free-form (within 24h service window), assign to agent, add label, update CRM, call webhook.
• Conditions — contact attribute, time of day, day of week, language, agent availability.
• Branching — multi-path with merge-back. Bots that escalate to humans and resume automatically when the human replies.
• Variables — pull from CRM, store, contact profile, conversation history, or external API.

How the pieces combine

A flow is just a sequence of these blocks. A simple, effective one might be: trigger on first message → condition on business hours → if open, send a welcome template and assign round-robin; if closed, send the after-hours template, capture category, and create a ticket. Each block is configured visually, and you can preview the path before publishing.

Where variables do the real work

Variables are what make a flow feel personal rather than robotic. Pulling first_name, order_status, or account_manager from your CRM means the same template reads differently for every customer. Because variables are resolved at send time, a template can serve thousands of conversations without manual editing.

Flow patterns compared

Why teams pick Greenbubble over a DIY WhatsApp chatbot

• Native templates — submit and reuse pre-approved message templates instead of building every reply from scratch.
• Multi-channel — same flow runs on WhatsApp, Instagram, Facebook, Telegram, and web chat.
• MCP-native automations — AI agents can read WhatsApp context and draft a reply, with a human approving before send.
• Real analytics — see flow drop-off, conversion, response time, all in one dashboard.
• Audit log — every action timestamped and attributed.

The human hand-off is the hard part

Most flow builders handle the first few steps well and then fail at the moment a customer needs a person. Greenbubble's flows escalate cleanly: the conversation moves to a named agent with the context captured so far, internal notes carry the transcript, and if the customer replies after hours the bot resumes without losing state. A flow that cannot hand off gracefully is worse than no flow at all.

Measuring whether a flow works

Track three numbers per flow: completion rate (how many contacts finish the intended path), deflection rate (how many are resolved without a human), and downstream conversion or CSAT. A welcome flow with a high completion rate but no lift in qualified leads is decoration. A flow that deflects 40% of routine questions and routes the rest accurately is doing real work.

Getting flow content right

• Keep the first message short. One idea, one question.
• Give tap options instead of asking people to type, wherever possible.
• Never pretend a bot is a human. Say "automated assistant" and let people reach a person.
• Respect quiet hours for anything that is not a reply to an inbound message.
• Make opt-out obvious and immediate.
• Test every branch before publishing, including the ones you think no one will take.

Pricing

Auto-reply and flow execution are included in every Greenbubble plan. You pay Meta's per-conversation fee for any business-initiated message and your normal Greenbubble SaaS fee. There is no per-flow or per-execution charge.

This is worth stating plainly because some platforms meter automation separately. If you run a flow ten thousand times, your cost is Meta's conversation fees plus the same flat SaaS fee — not ten thousand automation charges.

FAQ

Can WhatsApp auto-reply to customers?

Yes. Through the WhatsApp Business API, you can configure automatic greetings, away messages, and quick-reply buttons. For more sophisticated flows, Greenbubble's flow builder handles branching, CRM lookups, and human hand-off.

Do auto-replies count against WhatsApp conversation fees?

An auto-reply that is the first message from your business in a 24-hour window opens a new business-initiated conversation, which is billable at Meta's per-conversation rate. Welcome messages and certain utility categories may be free — see Meta's current pricing.

What is the difference between a chatbot and an auto-reply on WhatsApp?

An auto-reply is a single canned response triggered by an event (first contact, away hours, keyword). A chatbot is a multi-step flow that can branch, collect input, query APIs, and hand off to a human. Greenbubble supports both, and you can mix them.

Can I integrate my CRM with WhatsApp auto-replies?

Yes. Greenbubble's flow builder can look up a contact in your CRM (HubSpot, Zoho, Salesforce, Pipedrive), branch on their deal stage or lifecycle, and log the conversation back to the contact record.

Can an auto-reply send a discount code?

Yes. Greenbubble's flow builder can call a function to generate a unique discount code, attach it to a template, and send — all triggered by the customer's keyword or behaviour.

Will automating replies make my brand feel impersonal?

Not if the flows are written well. The failure mode is a bot that ignores what the customer said and repeats a menu. The fix is to keep tap options simple, capture the reason for contact early, and hand off to a person the moment the conversation leaves the script. Customers dislike bad automation, not automation itself.

Can a flow be triggered by something outside WhatsApp?

Yes. Triggers include webhook events and schedules, so your store, helpdesk, or internal system can start a WhatsApp conversation — for example, when an order ships or a subscription is about to renew. Greenbubble handles consent and template requirements so the send stays compliant.

How long does it take to build a first flow?

Most teams build a working welcome-and-route flow in an afternoon using the visual builder and the template library. The work is mostly in deciding what to ask and who should receive the hand-off, not in assembling the blocks.

Build your first WhatsApp flow today

Start free trial → · See flow templates → · Book a walkthrough →