What is predictive routing?

Predictive routing uses machine learning to rank each agent in your target agent pool for how well that agent is predicted to handle a specific interaction. In its simplest form, predictive routing identifies the agent-interaction match most likely to result in an optimal key performance indicator (KPI) value.

  • Predictive routing currently supports inbound interactions such as voice, email, and asynchronous messages. The message interactions include third-party messaging platforms, inbound SMS, Genesys Cloud web messaging, and open messaging.
  • Predictive routing is a routing method configured at the queue level, similar to bullseye or preferred agent routing.
  • You can use predictive routing on all queues, or only on selected queues where it has the strongest impact on a KPI.
  • Each queue using predictive routing can use a different KPI, depending on which KPI is optimal for that queue. Each queue optimizes only one KPI.
  • The list of currently-supported KPIs is available when you configure predictive routing. See Supported KPIs
  • Each KPI uses different data taken from various internal Genesys Cloud sources to predict the best agent for the current interaction. For a more detailed discussion of how predictive routing ranks agents and assigns interactions, see Agent identification process.

Start using predictive routing

There are three phases to activate predictive routing and take full advantage of all the benefits it can offer. The documentation for predictive routing fully explains each of these phases, listed below: 

    1. Benefit assessment – Evaluates your queues to determine whether they have high or low potential for optimization. The benefit assessment determines this potential for all available KPIs on each queue.
    2. Comparison test – Enables you to compare your existing routing method directly with predictive routing. Genesys Cloud routes interactions using the two methods alternately. You can view real-time and historical comparison test results for each queue you test.
    3. Enable ongoing value monitoring or activate full time:
      • Ongoing value monitoring – Enables you to route interactions using predictive routing 80% of the time interval and the baseline routing method 20% of the time interval. This increases the volume of predictive routing interactions, and allows for an increase in the benefits of predictive routing compared to a comparison test.
      • Active full time – In this phase, you have successfully proven the potential for benefit using predictive routing and use it to route all interactions on the queue.

To start using predictive routing, see the Use predictive routing section of Create and configure queues for instructions.

Supported KPIs

The following table lists the KPIs currently available for use in predictive routing. If a KPI you are looking to use is listed as Beta, contact your Genesys representative for assistance.

To submit a request for a new KPI, add it in the Genesys Cloud Product Ideas Lab, available to logged-in users from the Genesys Knowledge Network.

KPI Availability
Average Handle Time (AHT) Generally Available
Custom KPI Beta