Trigger strategies are referenced by the dataflow which is invoked by the channel
services. To test a trigger strategy, use the data transform (persona) that was automatically
generated for you based on your Context Dictionary and run the trigger strategy to identify
where the actions are dropping out or where the strategy results are unexpected.
Pega Customer Decision Hub
Before you begin: Update the persona data transform to include customer attributes
that satisfy any engagement policies that you defined for the strategy. The persona transform
includes the following sections:
ContainerPayload
Inbound channels use the real-time container service to execute the Next-Best-Action
strategy framework. This data structure allows you to define and test results for
inbound channels, such as the ContainerName,
Direction and Channel, as
well as any contextual data. To test Inbound scenarios, you need to set the Direction to
Inbound, and you must also set Channel to an active channel that you wish to test. For
Outbound scenarios you can set the Direction to be Outbound or blank; channel will be
ignored.
EventPayload
Real-time events execute the Next-Best-Action strategy framework to determine if there
is an appropriate next best action to take based on the event. This data structure
allows you to define and test results for an event such as
EventName and
EventType.
Customer, Account,
Subscriber, etc.
These sections represent all of the context dictionary related data. Add any property
that you want to test against your engagement policy criteria.
Note: The name of these
data structures will vary depending on your Context Dictionary definition.
JourneyStages
This section should be within the primary context entity section (for example,
Customer or Subscriber) and contains data about any active Journey Stages if the
Journeys feature is being used.
ActionInsights
This section should be within the primary context entity section (for example,
Customer or Subscriber) and contains data related to any Contact Policy Suppressions or
Arbitration Influencers.
Other sections may be present depending on your context dictionary configuration, and
sections may be added as required.
Save the persona data transform into your implementation
layer to create as many personas as you need to test your strategies.
In the Pega Customer Decision Hub portal, click IntelligenceStrategies.
Open the Trigger_NBA_TopLevel strategy.
Open the Test Run panel by clicking the slider button on the
upper right side of the strategy canvas.
In the Settings section, select Data
transform, and then select a persona transform, as in the following
figure:
Selecting a customer persona for strategy testing
Note: The example shows a configuration with three context levels. If your Context
Dictionary has only one or two context levels, your strategy will look simpler than
this.
Tip: When the trigger strategy is first generated, the strategy shapes may be
jumbled on the canvas. The simplest way to re-organize the shapes is to collapse each of
the context shapes – starting with the lowest level context – by clicking the collapse
button at the top left of the shape. You can then re-arrange the top-level shapes, and
then expand each context shape in turn and re-arrange the shapes within it.
Click the Open icon.
On the Definition tab of the data transform, specify the
characteristics of the persona.
The persona characteristics are defined as properties of the customer, their journey,
and other related data.
For example: If you have a specific test customer contact that you want to use as a test
persona, find the Set Action for the target
.CustID and enter the value into the
Source box. You can add new properties by right-clicking on a row
(not in the Target or Source boxes) and selecting Add
child.
Save the data transform (or click Save As to create a new
version) and switch back to your trigger strategy.
Click Run, and then select the Results
component to view the strategy results.
What to do next: To find the strategy component that is not performing as
expected, you can click through each shape on the strategy canvas and identify if and where
actions are dropping out or where the strategy results are unexpected.
