A customer asks about delivery times in the chat on your website. The next day they follow up by e-mail. A week later they write on WhatsApp that the parcel never arrived. If every channel runs a different bot with different knowledge, they get three different answers and your team gets three unconnected conversations in three systems. Omnichannel customer service is not about being everywhere. It is about saying the same thing everywhere.
One knowledge base is the foundation. Your delivery prices change on Monday. If the website chat, the e-mail autoresponder and the WhatsApp bot each have their own knowledge, someone has to remember three updates, and usually remembers one. For a week, customers in the forgotten channel get the old price. With a single knowledge base the change reaches every channel at once, and whoever owns the content has one place to keep right.
Same facts, different shape. Consistency does not mean identical replies. On WhatsApp the customer reads on a phone, so the answer should be short. By e-mail they expect a full explanation with a case number. In Teams, the person asking is often someone at a reseller who wants the fact without the pleasantries. The agent should adapt length and tone to the channel while drawing on the same approved sources and following the same rules: what it never promises, when it hands over to a person, which data it never asks for.
Shared analytics shows what one channel cannot. When conversations from every channel land in one place, you can see that the returns question comes up everywhere while invoice questions come mostly by e-mail. You can see that WhatsApp customers ask about order status because they write after buying, while website visitors ask which product to choose because they have not bought yet. With separate tools, those findings sit in three reports nobody puts side by side.
B2B channels and B2C channels. Teams and Slack suit business audiences: resellers, installers, partners, and your own staff looking for answers in internal procedures. These people spend their day there anyway, so they do not have to open another website. WhatsApp and Messenger are where consumers are, writing from a phone and often in the evening. E-mail stays for longer cases with attachments and a formal record. The website chat catches the first contact, before anyone leaves their details. Start with the channel where most questions arrive today, not the one that looks best in a slide.
What to watch out for:
- Identity. On WhatsApp you know a phone number, in Teams a company account, on the website often nothing. The agent should not share order details with someone it cannot link to a customer.
- Channel rules. WhatsApp lets a business reply freely for 24 hours after the customer's last message and only with an approved template after that, and the first message from a business requires opt-in. Every channel has rules like these, and you need to know them before launch.
- Human handoff. If each channel's conversations land in a different inbox, handoffs get lost. The person taking over should see the whole conversation and what the agent has already said.
- AI disclosure in every channel. From 2 August 2026, Article 50 of the EU AI Act requires that people interacting with an AI system are told so, at the latest at the first interaction. That applies to e-mail replies too, not only to chat.
- Sequence. Every channel at once means every problem at once. Launch two, improve the knowledge on real conversations, then add more.
When does one agent not make sense? When channels serve entirely different cases and people, for example technical support for installers and sales to consumers, with separate knowledge and separate teams. Two separately configured agents can then be clearer. Even so, it pays to bring the conversation insights into one place.
This is how we built Nexly Agents. One agent and one knowledge base cover the website chat widget, e-mail, WhatsApp, Facebook Messenger, Telegram, Signal, Microsoft Teams and Slack. Conversations from messaging apps land in one inbox with queues and a customer wait-time counter, and analytics agents read every conversation and report daily on the most common questions, missing information and lost sales. The agent detects the language of the conversation and handles 8 languages in one instance. More on channels: omnichannel AI agent for WhatsApp, Teams and Slack and AI agent for Microsoft Teams. If you are comparing tools, see our Intercom Fin alternative page.
Want to see this on your own website rather than on slides? Enter your website address and e-mail in the demo form. In 20 minutes you get an agent trained on your offer, ready to be quizzed.
Sources
- Meta for Developers, Send messages: customer service windows (24-hour window): https://developers.facebook.com/documentation/business-messaging/whatsapp/messages/send-messages
- Meta for Developers, Get opt-in for WhatsApp: https://developers.facebook.com/documentation/business-messaging/whatsapp/getting-opt-in
- Regulation (EU) 2024/1689 (AI Act), Article 50(1) and (5) and Article 113: transparency obligations applicable from 2 August 2026: https://eur-lex.europa.eu/eli/reg/2024/1689/oj