Deployment (Customer Insights)

Customer Insights Package is an extension of Sales Insights Package – it re-uses some Sales Insights components.

For the correct functionality, both Sales Insights Package1.6 and Customer Insights Package 1.6 (or later versions) have to be present at the partition.

In this section:

As the current version of PlatformManager does not support all of the steps needed to deploy the Customer Insights Package, the step Deploy & Run Calculation Flows (Manual Step) needs to be performed manually.

A. Delete Previously Created Objects

If you are going to run deployment to a partition where Customer Insights Package had been installed before, it is recommended to delete the previously created objects, namely:

  • Simulation (located in Analytics module > Simulation: CustomerInsights)

  • Company Parameters (main menu Company Parameters > CustomerInsights: all)

This step can be skipped if it is obvious that no changes were made to these objects (Company Parameters structure) since the previous version (see Customer Insights Release Notes).

B. Run Deployment from PlatformManager

  1. Get access to PlatformManager and target partition, as described in common installation prerequisites.

  2. Access PlatformManager at https://platform.pricefx.com/ and log in.

  3. Go to Marketplace and find the Customer Insights package.

  4. Click the accelerator package tile, select the partition where you want to deploy the package and confirm the deployment dialog to start.
    For detailed description of all deployment options, see PlatformManager documentation.

  5. Set up Datamart mapping.

    Select the Datamart name and field names (from the Datamart) for mapping and click Continue.

  6. Wait for the system to finish the import.

  7. The deployment is complete.

C. Update Price Parameters and Advanced Configuration Options

The PFXTemplate_CustomerInsights_Customer_Segment Company Parameter needs to be consistent with the partition data for the fields set as Active. This CP defines the fields used for a customer segmentation.

If the partition has some special setting (different from the Accelerator default setting), check and update the settings in other Company Parameters.

For more details see Customer Insights configuration stored in Company Parameters.

D. Verify Number of Threads at Partition

Starting with Customer Insights Package version 1.6, it is recommended to have 8 threads per partition where Customer Insights is deployed. If the number is lower, Data Load calculations may fail.

By default, the value is lower, so you need to check with Pricefx Support. Threads are set by them as a server config parameter numberOfThreadsPerPartition for both paQuery and paDataLoad.

E. Configure Batching

Batching for the aggregated Data Load must be configured before the calculation is triggered (next step). For more details, refer to https://pricefx.atlassian.net/wiki/spaces/ACC/pages/5386929458/Data+Flow+Customer+Insights#Configurable-Batching-for-Aggregation-Data-Load.

F. Deploy & Run Calculation Flows (Manual Step)

To keep the aggregated data (CI_AggregatedData Data Source) synchronized with customer classification data (CI_CustomerClassification Data Source), a scheduled job needs to be set to run the Data Load CI_CustomerClassification first and then run CI_AggregatedData Data Load (Distributed Calculation).

This can be done in Calculation Flows by taking these steps:

  1. Go to Administration > Configuration > Calculation Flows.

  2. Find a calculation flow with the name “CustomerInsights_Sequencer”.

     

  3. Select the “CustomerInsights_Sequencer” and update Time Unit and Number of Time Units.
    You define here how often the data will be refreshed and when. This setting has a big impact on the dashboard results and their validity.

     

  4. Click the Deploy button to deploy the calculation flow.

     

  5. When finished, the system will show more lines for the flows.

     

  6. If you want to run the calculation flow immediately, select the row in the right window and then click Run now.