How-to
Article MarkdownReport an error

How to notify the client when an operator does not reply for a long time

The operator is already in the chat, but the answer takes time: they check details with colleagues or wait for data from another system. The client cannot see that the request is in progress and writes again. The operator idle phase of the Operator queue cascade node handles this for you. It sends the client your message when the operator has not replied within the time you set.

When to use it

  • Operators often need to check information, so a dialog pauses for 15–30 minutes.
  • Clients ask "Are you there?" while the operator prepares an answer.
  • The client should know that the request is in progress even when the operator stays silent for a long time.

What you need to know

  • Operator idle phase — the part of the Operator queue cascade node that works after an operator joins the chat. You set it up in the Operator idle section. The silence countdown starts when the operator joins.
  • Public reply — an operator message the client sees. While the phase is running, such a reply restarts the countdown from the first step.
  • Internal note — an operator message that clients do not see. Internal notes and client messages do not change the countdown.
  • Names in the node form — section titles, field names, and Operation options are shown in Ukrainian in any interface language. This guide gives the English name first and the form text in «», for example Lifecycle («Життєвий цикл»). Step settings, modes after the last step, and form warnings follow your interface language, so they appear without «».

Before you start

Steps

1. Add the cascade node after Connect to Operator

  1. Open the scenario in Scenario Builder.
  2. Add an Action node: in the Add Action window, select the Operator queue cascade («Каскад черги оператора») template — you can find it with the search. If the scenario already has a queue cascade node, use it.
  3. Connect these outputs of Connect to Operator to the cascade node:
    • success;
    • queue and connected, if the node has these outputs.
  4. Click the cascade node to open Node Inspector.

The cascade node must run in every dialog where a chat is created. If only the queue output leads to it, the phase does not work for chats that an operator joins immediately. If the scenario continues differently after these outputs, add a separate cascade node with the same settings to each branch.

Important: Do not place the cascade node before Connect to Operator. Otherwise, the phase does not start for a client who contacts you again.

2. Check lifecycle

  1. In Lifecycle («Життєвий цикл»), set Operation («Операція») to Start cascade («Запустити каскад»).
  2. Keep Stop when operator joins («Зупиняти при підключенні оператора») and Stop when chat is closed («Зупиняти при закритті чату») enabled. The phase does not work without the first switch.

If the client does not need queue reminders, do not add steps in Reminder steps («Кроки нагадувань») — the node then runs only the operator idle phase.

3. Set up the message for the client

  1. Open the Operator idle («Оператор мовчить») section.
  2. Enable Auto-message when a connected operator does not reply («Автоповідомлення, коли підключений оператор не відповідає»).
  3. In Operator idle steps («Кроки фази «Оператор мовчить»»), click Add step.
  4. In Delay after previous step, enter the silence time, for example 30, and in Unit select Minutes. For the first step, this time counts from the moment the operator joins. Here, the Approx. from cascade start hint under the field also shows the time since the operator joined.
  5. Enable Send message and fill in Message.
  6. In Transition, select No transition. A new step is created with Event branch, so change it manually.

If the client should get several messages, add more steps. Each following step runs after its own delay, counted from the previous step.

4. Choose what happens after the last step

  1. In After the last idle step («Після останнього кроку фази»), select a mode:
    • Stop cascade — the phase ends after the last step. Even if the operator goes silent again, no new messages are sent.
    • Repeat last step — the last message repeats while the operator stays silent. After the operator replies, the countdown starts from the first step.
  2. If you selected a repeat, the Repeat delay field appears — set the repeat interval there. If you leave 0, the step repeats after its own delay.

For the Repeat from first step mode, see Operator queue cascade parameters.

5. Publish the scenario

Click Publish. The settings apply to new dialogs right after publishing.

What happens next

Suppose the step is set to 30 minutes. The operator joins at 10:00 and writes "Hello, checking now" at 10:05. If the operator stays silent after that, the client gets your message at 10:35.

If the operator writes again at 10:30, the countdown starts over, and the message arrives no earlier than 11:00. In the Stop cascade mode, the phase ends after this message.

While the phase is running, the countdown also reacts to other chat events:

  • the chat is closed — no message is sent;
  • the chat is transferred to another operator — the countdown starts over;
  • the operator leaves the chat without closing it, or the chat is returned to the queue — the phase waits for the next operator to join.

Check the result

  1. Temporarily set the first idle step to 20 with Seconds as the unit and publish the scenario.
  2. Start a test dialog as a client. Join the chat as an operator or wait for automatic distribution to assign it.
  3. Do not reply to the client for 20 seconds — the message should arrive.
  4. Close the chat and repeat step 2 with the same client. Do not reply for 20 seconds — the message should arrive again.
  5. In the next test dialog, reply to the client 10 seconds after joining. The message should arrive 20 seconds after your reply.
  6. Restore the working delay and publish the scenario again.

If the message does not arrive

  • Check that Auto-message when a connected operator does not reply is enabled and the scenario is published.
  • Check that the cascade node sits after Connect to Operator and is connected to the output the dialog took: success, queue, or connected.
  • If the first message arrived but later ones did not, check the mode in After the last idle step. In the Stop cascade mode, the phase ends after the last step.
  • If everything is set up but no messages arrive, contact your platform administrator: the operator idle feature must be enabled for your platform.

Common warnings

Warning What to do
Operator idle phase is enabled, but it has no steps. Add a step in Operator idle steps or turn the phase off.
Operator idle phase works only together with «Stop when operator connects». Enable Stop when operator joins («Зупиняти при підключенні оператора») in Lifecycle («Життєвий цикл»).
Idle step 1: event branch is selected, but branch ID is empty. In Transition, select No transition or fill in Branch ID.
Idle step 1: send message is enabled, but the message text is empty. Fill in Message or turn off Send message.
Repeat after the last idle step has zero delay… Set an interval in Repeat delay, or make sure the last step's delay is above zero: the step then repeats after that delay.

See also

Using your own AI agent?Install the open skill so your AI agent can work with current official ConnectiveOne documentation.Get the skillNeed support?Couldn’t find the answer or need help from the ConnectiveOne team? Create a request in Client Portal.Create a request