Omnichannel chatbots: one AI brain for WhatsApp, Instagram, web and calls
An omnichannel chatbot uses one AI agent, knowledge base and set of flows across every channel a customer might use — WhatsApp, Instagram DMs, website chat and phone calls — so answers stay consistent no matter where someone reaches out, and your team manages every conversation from a single inbox instead of separate tools per channel.
- One knowledge base and system prompt can power the same AI agent across WhatsApp, Instagram, web chat and voice.
- Each channel still needs its own channel-specific blocks — buttons work differently on web widget vs WhatsApp, for instance.
- A shared team inbox means your staff don't need separate logins for each channel's native tool.
- Customer identity (phone number, name) can carry across channels where the channel provides it, keeping context consistent.
- Start with your two highest-volume channels rather than launching on every channel at once.
Why one brain beats separate bots per channel
Many businesses end up with a WhatsApp bot built one way, an Instagram auto-reply built another way, and a website chat widget answering differently again, because each channel was set up separately, often at different times. The result is inconsistent answers and duplicated work every time a policy changes.
An omnichannel chatbot uses the same AI agent, same system prompt, same knowledge base, across WhatsApp, Instagram DMs, Facebook Messenger, Telegram, the website chat widget, SMS and voice calls. Update your product pricing once, and every channel answers with the same information. This is different from an omnichannel inbox alone, which only unifies where you read messages: the brain behind the replies needs to be shared too.
What stays the same, what changes per channel
| Element | Shared across channels | Channel-specific |
|---|---|---|
| Knowledge base / FAQs | Yes | No |
| System prompt / tone | Yes | No |
| Buttons / List Menu / Carousel | No | Yes, rendered differently per channel |
| WhatsApp Flow (native forms) | No | WhatsApp only |
| Voice blocks (gather, DTMF, transfer) | No | Voice calls only |
| Team inbox | Yes, one shared view | No |
Understanding this split matters when you're building: reuse the knowledge base and system prompt everywhere, but expect to rebuild the interactive parts of a flow (buttons, forms, menus) once per channel, since each channel's interface works differently.
Setting up an omnichannel flow on Convo360
- Create one AI agent with a clear system prompt and upload your knowledge base documents once, see how to train an AI agent on a knowledge base.
- Connect each channel you need: WhatsApp (embedded signup), Instagram/Facebook (Meta connection), the web chat widget (embed code), and voice (telephony provider).
- Build a core flow with the shared logic, then adapt channel-specific blocks: a Carousel for WhatsApp product browsing, a Show Form block for web enquiries.
- Route all channels into the same live-chat team inbox so your team sees every conversation in one place, with human take-over pausing the bot automatically.
- Check per-channel Analytics to see where most conversations actually happen.
What it costs to run channels together
Adding channels is mostly a plan and configuration decision, not a per-channel fee. On the Starter plan (₹999/month, ex-GST) you get one workspace with WhatsApp and web chat; the Pro plan (₹2,999/month) adds AI voice calls and every integration across three workspaces; Business (₹7,999/month) removes the workspace limit and adds the REST API. Whichever plan you're on, WhatsApp conversations, AI tokens and call minutes are billed directly by Meta, your AI provider and your telephony provider at their own rates. Convo360 doesn't add a per-channel markup. See WhatsApp chatbot pricing in India for the fuller breakdown.
Keeping identity consistent across channels
Where a channel provides identifying information, a phone number on WhatsApp, a caller ID on voice, Convo360 can match it to the same contact record, so a customer who messaged on WhatsApp yesterday and calls today doesn't have to repeat their order number. Use system variables like {{phone}} and custom fields in Contacts to carry this context between flows.
Web chat and anonymous Instagram interactions are harder to tie to an existing contact until the person shares identifying details. Plan your flow to ask for a phone number or email early if continuity across channels matters for that use case.
Mistakes to avoid when going omnichannel
- Launching every channel on day one instead of proving the flow on your busiest channel first.
- Writing a separate, inconsistent knowledge base per channel instead of one shared source of truth.
- Forgetting that a WhatsApp-specific block like WhatsApp Flow simply won't exist on the website widget, and needs a channel-appropriate alternative like Show Form.
- Not routing every channel into one team inbox, leaving staff checking multiple logins anyway.
- Ignoring per-channel analytics, so you keep investing effort in a channel your customers barely use.
Where to start, and a quick readiness checklist
Don't launch on every channel simultaneously. Start with your two highest-volume channels, for most Indian SMEs that's WhatsApp plus either Instagram or the website, get the flow and knowledge base right there, then extend to additional channels once it's working well. See Integrations docs and Instagram & Messenger docs for channel-specific setup.
Going omnichannel is worth doing once, not once per channel. Check these first:
- Do you already get meaningful enquiry volume on at least two channels, not just WhatsApp alone?
- Is your knowledge base (product info, pricing, policies) written down and current, ready to upload once?
- Does your team currently juggle multiple app logins, WhatsApp, Instagram, a separate web chat tool, that a shared inbox would simplify?
- Do you need voice calling too, or is chat-only enough for now?
- Have you picked which plan tier (Starter, Pro or Business) matches the channels and workspace count you need?
If most answers are yes, an omnichannel setup will save your team real time within the first month rather than adding complexity. If you're still only active on one channel, it's usually better to get that one right first, see the WhatsApp chatbot guide for businesses in India as a starting point, or AI agent vs chatbot to understand what you're actually building before expanding channels.
Example conversation
Connect all your channels free for 14 days
Start free in minutes — no code, no credit card. Bring your own AI keys and pay no per-message markup.
Frequently asked questions
Do all channels need the same flow logic in an omnichannel chatbot?
The core knowledge base and answers should stay consistent, but each channel typically needs its own channel-specific blocks — like WhatsApp Flows for native forms or a Show Form block for the web widget — since interface capabilities differ.
Can I manage WhatsApp, Instagram and web chat from one inbox?
Yes, Convo360's live-chat team inbox brings conversations from every connected channel into a single view, so your team doesn't need separate logins per channel, and human take-over pauses the AI agent automatically whenever a staff member replies.
Does an omnichannel chatbot know if the same customer contacted me on two channels?
Where a channel provides identifying details like a phone number, Convo360 can match it to an existing contact record; fully anonymous channels need the customer to share identifying details first.
Should a small business start with all channels or just one?
Start with your one or two highest-volume channels, get the flow and knowledge base working well there, then expand to additional channels once the core setup is proven.
Does voice calling share the same AI knowledge base as WhatsApp?
Yes, the same AI agent and knowledge base can power both chat and voice, though voice flows use their own blocks (gather, DTMF, transfer) suited to a phone conversation.
Related reading
Available across India — see WhatsApp & AI calling automation by state and city.