How can we help you?

Smart Routing and Idle Escalation

Smart routing keeps conversations moving even when an agent steps away mid-chat. It combines an initial response timeout, a secondary idle check after the agent has been talking, agent alerts, automatic re-routing to another available agent, and a configurable fallback when no one is available.

Everything below is governed by an Auto-Routing Policy that you can set per site and override per team. Velaro ships with sensible defaults, so smart routing works out of the box without any setup — adjust it only if you want different behavior.

The two timers

1. Initial response timeout. When an agent is assigned a conversation but never responds, Velaro warns the agent and then routes the conversation to the next available agent. This protects visitors from an agent who accepted a chat and went quiet.

2. Idle-after-talk escalation. Once an agent has responded, a second idle timer starts. If the agent then goes silent while the visitor is waiting:

  • The agent receives a clickable alert ("This chat is waiting on you — still there?"). One click keeps the conversation.
  • If the agent does not acknowledge within the grace period, Velaro escalates the conversation to the next available agent.
  • If no agent is available, Velaro applies the policy's terminal action (see below) and flags the conversation so the original agent is reminded to follow up later.

When no agent is available

If everyone is busy or offline, the conversation does not silently stall. Velaro applies the terminal action defined in your policy. The default is to send the visitor a post-chat survey and mark the conversation for follow-up, so the original agent can reach back out once they are free. Other terminal actions you can choose include keeping the visitor waiting, transferring to another team, or sending an offline message.

No bouncing in circles

A conversation is never routed back to an agent who already idled out of it. Velaro tracks which agents have had a turn, and once every eligible agent has been tried, it logs an "all agents tried" event and applies the terminal action. The "will be re-routed" warning is suppressed when there is no available agent to route to, so agents are not told a chat will move when it cannot.

The Auto-Routing Policy

You configure smart routing through an Auto-Routing Policy. Set a site-wide default and override it per team. Available settings:

  • Enabled — turn smart routing on or off.
  • Initial response timeout — how long an assigned agent has to send a first reply.
  • Idle-after-response timeout — how long an agent may be silent after replying before the idle alert fires.
  • Idle alert grace — how long the agent has to acknowledge the alert before the chat escalates.
  • Re-route attempts — cap the number of agents tried, or keep going until everyone has had a turn.
  • Terminal action — what happens when no agent is available (default: Survey).
  • Agent follow-up flag — mark unhandled conversations so the original agent can follow up (on by default).
  • Message templates — the agent alert and visitor-facing messages, all editable.

Defaults, so you don't have to set anything up

Every new site inherits a built-in default policy: smart routing on, reasonable timeouts, survey as the terminal action, and follow-up flagging enabled. You only need to open the policy if you want to change something. Each team can override the site default without affecting other teams.

Reporting

Every step of an escalation is recorded — warned, acknowledged, re-routed, exhausted, and terminal action applied — so you can see exactly how conversations were handled and tune your timeouts over time.

Share: Email

Was this article helpful?