WhatsApp CRM on the WhatsApp Business Platform

This guide is for a CTO, product owner or operations lead adding WhatsApp as a two-way, CRM-synced channel to an AI receptionist, support desk or sales pipeline, and deciding whether to buy a SaaS WhatsApp inbox, build on the Cloud API directly, or use a product Netbase already operates. For the delivery work around it, see systems and API integration and custom ERP and CRM development; for the architecture this channel slots into, see the enterprise systems integration guide.

In this guide

• What is WhatsApp CRM on the WhatsApp Business Platform?
• What it does
• What architecture does it use?
• Use cases
• Where AI fits
• How Netbase delivers it
• Alternatives: buy a SaaS WhatsApp inbox, build on the Cloud API, or use WorkChat
• Limits of this guide
• Frequently asked questions
• Next step

What is WhatsApp CRM on the WhatsApp Business Platform?

The WhatsApp Business Platform is Meta's official API (the Cloud API) for sending and receiving WhatsApp messages programmatically, distinct from the free WhatsApp Business app and from third-party inbox tools that sit on top of it. Putting it "in the CRM" means a company's own systems, not a shared vendor inbox, hold the conversation: messages sent and received through the Cloud API are matched to a contact record, logged on that contact's timeline, and available to the AI-reply and hand-off logic the CRM runs. Netbase builds this layer directly on the Cloud API: Meta-approved message templates by category, opt-in broadcasts with segmentation, delivery and read status delivered through webhooks verified with the X-Hub-Signature-256 HMAC signature, idempotent sends so a retried request never double-sends, and AI-drafted and automatic replies for common questions with a human hand-off, all synced into the CRM (F193).

What it does

• Sends Meta-approved message templates by category — marketing, utility and authentication — each submitted for Meta's review before use outside an open conversation.
• Runs opt-in broadcasts with tags and segments, paced to protect the WhatsApp number's quality rating rather than sent as one uncontrolled blast.
• Tracks delivery and read status per message through Meta's webhooks, with every inbound notification verified against its X-Hub-Signature-256 signature before the CRM acts on it.
• Sends with idempotency keys, so a retried API call from the CRM or the AI receptionist never results in a duplicate message reaching the customer.
• Handles the 24-hour customer-service window: free-form replies inside the window, an approved template required to reach a customer outside it.
• Drafts AI replies and sends automatic replies to common questions inside the window, with a clear hand-off to a human agent for anything the AI should not resolve alone.
• Syncs contacts, conversation history and notes into the CRM, so a WhatsApp thread reads like any other channel on the customer's record rather than living only on a phone.

Use cases

• AI receptionist on WhatsApp. A model-agnostic AI assistant (Netbase works with the major commercial and open AI tools and models, chosen per project, F070) answers common questions and books or confirms appointments inside the service window, handing off anything it should not resolve alone.
• Transactional and utility messages. Order confirmations, appointment reminders and account alerts sent as utility templates, matched to the triggering CRM event.
• Opt-in broadcast campaigns. Marketing-template sends to a tagged, opted-in segment, paced to protect the number's quality rating rather than sent as a single uncontrolled blast.
• Two-way support synced to the CRM. Inbound replies land on the contact's timeline instead of a phone inbox, so a human agent or the next automated step can act on them.

Where AI fits

• AI-drafted replies run inside the 24-hour customer-service window, where free-form messages are allowed, and stop being a safe default the moment a conversation needs a template to continue.
• Automatic replies answer common, repeatable questions (hours, order status, booking availability); anything outside that scope hands off to a human agent rather than guessing.
• The AI layer is model-agnostic: Netbase works with the major commercial and open AI tools and models, choosing per project rather than committing to one vendor's stack (F070).

How Netbase delivers it

Netbase scopes the WhatsApp channel with the client first: which CRM or AI receptionist it must sync into, which template categories the business needs (utility, marketing, authentication), and whether the client wants its own Meta Business Account and Cloud API credentials or Netbase's own Cloodo Business Division to operate the channel through its WorkChat product, a WhatsApp Business API and AI chat product in Netbase's productized module library (F025, F024). The build adds the send service with idempotency keys, the signed-webhook receiver, template submission and category management, broadcast segmentation and pacing, and the AI-reply logic with its hand-off rule, then syncs every event into the CRM. Hand-over includes the API and webhook documentation the client's developers need, and Netbase can also build and maintain the CRM or AI-receptionist layer that sits on top through custom ERP and CRM development and systems and API integration. Most Netbase projects are delivered on fixed-price contracts, with a monthly team retainer or milestone-based model also offered (F067). This channel work sits alongside Netbase's anonymised delivered project combining an AI chatbot with CRM sync (see WhatsApp AI chatbot CRM integration), and pairs with Netbase's Android SMS gateway as the fallback channel for customer numbers that do not have WhatsApp installed.

Alternatives: buy a SaaS WhatsApp inbox, build on the Cloud API, or use WorkChat

Limits of this guide

• Meta's template categorization, category-based pricing and webhook policies change; verify the current rules at Meta's developer documentation before planning a rollout (sources below).
• This page describes a capability Netbase builds and runs; it does not name a client, a message volume, a price or a Meta partnership status, because none of those is a registered fact for this capability (F193).

Next step

Netbase scopes a WhatsApp CRM build around the CRM or AI receptionist it needs to plug into; talk to the CEO about your integration to review template categories, broadcast segments and the backend it should connect to.

Related services and solutions