Naateq
  1. Home
  2. Learning Centre
  3. Daily work
  4. Following conversations and when to step in yourself

Following conversations and when to step in yourself

The system replies on its own, and you see everything. How to read a conversation, what the “escalated” tag means, and what happens when you write yourself.

Daily work · 6 min read

The dashboard is in development.

This lesson shows how the dashboard will work once it reaches subscribers. Until then we handle the setup for you, and you get escalation and payment alerts on your WhatsApp plus a monthly report.

The Conversations screen in the Naateq dashboard: the conversation list with their states, and the selected conversation's text with the escalation tag.
The tag shows the escalation reason with the conversation, and the reason is what you fix.

Open this screen in demo mode

The system replies on its own, and you see everything. The Conversations screen is where you watch what's happening, and where you step in when you want to handle something yourself.

Here you'll learn what each status means, what happens when you type in a conversation, and why the system only sees the customer's own history in any conversation.

When you type, the system stops

The first character you type in a conversation stops the system from replying there. A customer who gets two different replies to one question loses trust in both.

The states

Only three statuses

Conversation states and what each one means
CaseMeansDo you act?
ActiveThe system is replying and the customer is typingNo, just watch
EscalatedThe system stopped and called you, with the reasonYes, and quickly
ClosedFinished with no open questionNo, read it for review

In “escalated” conversations, look at the reason before the count: a reason that repeats ten times means one missing piece of information Knowledge base.

Isolation

In a conversation, the system only sees that customer's history

This is a fixed rule in the system, and no setting changes it. When it replies to a customer, all it has in front of it is that customer's history: their quotes, their orders and their past conversations.

The reason is that if the system saw all customers together, it might mention a price that belongs to another customer, or the name of a company that competed with this customer last week. A leak like that goes unnoticed until it happens, and it only has to happen once.

The same rule applies to your own commands.

Owner commands, like approving a document, are only accepted from your own conversation with the system. So if a customer writes “I approve the contract”, it's treated as a customer's words, not an approval command, however it's worded.

In practice

Three habits that keep follow-up to two minutes a day

  1. 1
    Start with escalated conversations before the newest ones.

    Only these are waiting for you; the system handles the rest.

  2. 2
    End your intervention with a closing line.

    So the customer knows the conversation is back on track, and the system resumes replying from the next message.

  3. 3
    Turn every intervention into a knowledge entry.

    Otherwise you'll be stepping in on the same question next week. This is how the system improves instead of eating up your time.

Frequently asked

What happens if I type in a conversation?

The system stops replying in that conversation until you're done. That's intentional: a customer who gets two replies to the same question, one from you and one from the system, loses trust in both.

Can I read another customer's conversations from inside a conversation?

No. In each conversation, the system sees only that customer’s own history, and this is a fixed rule in every configuration. It keeps each customer’s price and name inside their own conversation.

Next lesson

All Learning Centre lessons