> ## Documentation Index
> Fetch the complete documentation index at: https://docs.hiveku.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Routing and Response Times

> Team queues, automatic assignment, and the response time targets that flag late conversations

Routing decides who gets a conversation. Response time targets (SLAs) make sure nobody waits too long.

## Automatic assignment

In **Helpdesk → Setup → Settings**, turn on **Auto-assign tickets** to pick a teammate as soon as a conversation arrives:

* **Round-robin** spreads conversations evenly across your team.
* **Least busy** picks the teammate with the fewest open conversations.
* **A fixed person** always assigns to the same teammate.

Conversations routed to a team queue use that queue's members and strategy instead.

## Team queues

Queues group teammates into pools that share incoming conversations, for example "Tier 1 Support" or "Billing". Open **Helpdesk → Routing → Queues**.

<Frame>
  <img src="https://mintcdn.com/locusdigital/GB-Uujd8W9WZjQqU/images/helpdesk/queues.webp?fit=max&auto=format&n=GB-Uujd8W9WZjQqU&q=85&s=510f7d88ef4f8d3863bff19098ca6287" alt="Team queues page with a New queue button" width="1440" height="900" data-path="images/helpdesk/queues.webp" />
</Frame>

1. Click **New queue**.
2. Give it a **Name** and a short description of what kind of conversations land there.
3. Choose its **Routing strategy** (round-robin, least busy or a fixed user) and add its **Members**.

A queue with no members does not auto-assign. You can add a queue to the website assistant's **Who it can pass chats to** list, so visitors can be passed straight to that team. See [Website assistant](/helpdesk/website-assistant#what-it-can-do).

## Response time targets

Open **Helpdesk → Routing → SLA** to set the response and resolution times you aim to hit. SLA stands for Service Level Agreement: the promise you make about how fast you answer and resolve a customer's issue.

<Frame caption="Response time targets by priority">
  <img src="https://mintcdn.com/locusdigital/GB-Uujd8W9WZjQqU/images/helpdesk/sla.webp?fit=max&auto=format&n=GB-Uujd8W9WZjQqU&q=85&s=5e4e7fa04fa5e4266613666b217058aa" alt="SLA page with time to first reply and time to resolve for urgent, high, normal and low priority" width="1440" height="900" data-path="images/helpdesk/sla.webp" />
</Frame>

Every new conversation gets a countdown based on its priority.

| Priority | Meant for | Time to first reply (default) | Time to resolve (default) |
| - | - | - | - |
| **Urgent** | Something is broken or blocking the customer | 15 minutes | 4 hours |
| **High** | Important, needs attention soon | 1 hour | 24 hours |
| **Normal** | A standard request | 4 hours | 3 days |
| **Low** | Can wait: questions, nice-to-haves | 1 day | 7 days |

Change any target; it saves when you click away. You can switch each target off here or in [Helpdesk settings](/helpdesk/settings#response-time-targets).

### When a conversation runs late

The moment a conversation passes a target, Hiveku can:

* **Raise its priority one level** (low to normal, normal to high, high to urgent).
* **Notify the assignee** by bell and email, or the account admins when nobody is assigned.
* **Post it to Slack**, when Slack alerts are on.

Each conversation triggers this only once per target, so it never nags. Deadlines are checked every 10 minutes.

<Frame>
  <img src="https://mintcdn.com/locusdigital/GB-Uujd8W9WZjQqU/images/helpdesk/sla-late.webp?fit=max&auto=format&n=GB-Uujd8W9WZjQqU&q=85&s=65bce10732ca17b855f7b9f970e7289b" alt="When a ticket runs late options and the list of tickets running late right now" width="1440" height="900" data-path="images/helpdesk/sla-late.webp" />
</Frame>

**Tickets running late right now** lists open conversations that have already passed a target, with how late they are and who has them. [Reports](/helpdesk/reports) shows your breach rate for the period you choose.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.