Amazon Connect – Does queue delay on a higher priority queue block lower priority queues?

Question

I’m trying to understand the exact interaction between queue priority and delay in Amazon Connect routing profiles, as the documentation examples don’t explicitly cover this scenario.

Scenario:

Queue A → Priority 1, Delay 30 seconds
Queue B → Priority 2, Delay 0 seconds
Single available agent
One contact in each queue

Will the agent be offered the Queue B contact immediately or will the presence of a higher-priority contact in Queue A (even though still in delay) prevent Queue B from being routed, potentially leaving the agent idle?

In other words, do contacts that are still within their queue delay period participate in queue prioritisation decisions, or are only contacts eligible to be offered considered?

I’m specifically looking to confirm actual system behaviour, as this has implications for multi-queue routing design.

re:Post Agent · Accepted Answer

Based on how Amazon Connect routing profiles work, contacts that are still within their delay period do NOT block lower priority queues from being routed. The delay mechanism holds contacts for the specified period even when agents are available, but this doesn't prevent lower priority queues from being processed.

In your specific scenario with Queue A (Priority 1, Delay 30 seconds) and Queue B (Priority 2, Delay 0 seconds), the agent would be offered the Queue B contact immediately. Here's why:

The priority system in Amazon Connect works by ensuring all higher priority calls must be processed before lower priority calls will be processed. However, a contact in Queue A that is still within its 30-second delay period is not yet eligible to be processed - it's being held and cannot be routed to agents during that delay window.

Since the Queue A contact isn't eligible for routing yet (due to the delay), and Queue B has no delay, the Queue B contact becomes the highest priority eligible contact and will be routed to the available agent immediately.

The delay feature is specifically designed to reserve certain agents for escalation or SLA management scenarios. Contacts are held for the delay period even when agents are available, allowing lower priority work to be handled first. Only after the delay period elapses does the contact become eligible and participate in the priority-based routing decisions.

So in practical terms: only contacts that have passed their delay period are considered eligible for routing and participate in queue prioritization decisions.  
**Sources**  
[Operational excellence in Amazon Connect workloads - Amazon Connect](https://docs.aws.amazon.com/connect/latest/adminguide/operational-excellence.html)  
[Queue priority and delay examples to help you load balance Amazon Connect contacts - Amazon Connect](https://docs.aws.amazon.com/connect/latest/adminguide/concepts-routing-profiles-priority.html)

Clarence Choi · Answer

To add the explanation, if you have calls in Queue A that are over 30 sec wait time.  And you have calls in Queue B less than 10 sec as example.  In your configuration, if an agent belong to this Routing Profile will get a call from Queue A before Queue B,  in fact, all agents in in this RP will clear all calls that are over 30sec before calls from Queue B will be routed to them, makes it look like Queue A is blocking Queue B as long as there is a call waited >30sec.

This is very unusual setting between these two queue in RP.  What exactly is your use case?  Normally without using Proficiency Routing, you would set something like this:

A - Pri 1, Del 0
B - Pri 1, Del 30

if there are calls over 30 sec in B, and calls in A, all calls in A will be eligible for routing, and all calls over 30 sec in B will be eligible.  As Pri are the same, longest call among the above will be routed to next agent.  So if there are calls waited in A for 5 min, and calls in B for 1 min.  Calls in A will be picked up since it wait longer, so calls in B will not block A as Pri is the same.

Another thing to note is

A - Pri 2, Del 30
B - Pri 1, Del 0

In here, agents will clear queue B first then help calls in A that are over 30 sec wait time.  A will block B but that is by design.  The design here is agents in RP is focus on B, and if and only if they have capacity, they will help in A.  And even if there are calls waiting in A for 10 min, and if there any call just come into B, next available agent will take B (2 sec wait) before A (10 min wait)

Now if you put the TWO routing profile together.  You are now covered for both queue as long as you have agents in both.  On an agent surplus in both queue, all calls will be handled immediately, no block each other.  So as long as you have a Routing Profile that covers a queue's pri and del to be Pri 1 and delay 0.  You should be ok, so listing all your Routing Profile across is actually more important understand if blocking scenario will occur.

Amazon Connect – Does queue delay on a higher priority queue block lower priority queues?

Add your answer

Relevant content