Testing Next-Best-Action configuration with audience simulations
You can improve the performance of your Next-Best-Action Strategy by testing the configuration of engagement policies against simulated audiences. In this way, you can check how many potential actions are filtered out by each component of the policy, and discover if a particular criterion is too broad or too narrow for your requirements.
On the Engagement policy tab, click the name of any group, or click All groups to run the simulation for all issues and groups in your business structure.
In the Audience simulation section, select or create a simulation with which you want to test the proposition filter:
If you are creating your simulation for Scenario Planner purposes, you must select Engagement policies and arbitration as the Next-best-Action scope. Only the simulations run with arbitration will work with Scenario Planner.
- To use an existing simulation, select the simulation in the Simulation list.
- To create a new simulation, click Create simulation, and then select the audience to use for the simulation.
After the simulation test finishes, analyze the results to see what percentage of the audience would receive each action according to the current engagement policies.For each customer context in your application, the simulation drills down to individual components of the engagement policy. The results show numerical or percentage values that indicate the population of the selected audience that receives the proposition based on the current criteria. For example, if the result for a criterion that checks if the proposition is active returns 100.00%, the condition did not filter out any audience members.
- Seeing the effect of your business changes with Scenario Planner
Understand the effect of changes to your Next-Best-Action Designer configuration by using the planning capability of Pega Customer Decision Hub. Use scenario planning to examine the impact of adding new actions, changing lever values, or modifying next-best-action engagement policies.