As each message is sent to the router, the routees are randomly ordered. The message is sent to the first routee. If no response is received before the interval has passed, the same message is sent to the next routee. This process repeats until either a response is received from some routee, the routees in the pool are exhausted, or the within duration has passed since the first send. If no routee sends a response in time, a akka.actor.Status.Failure wrapping a akka.pattern.AskTimeoutException is sent to the sender.

The goal of this routing algorithm is to decrease tail latencies ("chop off the tail latency") in situations where multiple routees can perform the same piece of work, and where a routee may occasionally respond more slowly than expected. In this case, sending the same work request (also known as a "backup request") to another actor results in decreased response time - because it's less probable that multiple actors are under heavy load simultaneously. This technique is explained in depth in Jeff Dean's presentation on Achieving Rapid Response Times in Large Online Services.