Project Recipe: Scoring for Salesforce

The following is a guideline to follow when using the recipe "Scoring for Salesforce." Please read these guidelines before cloning the recipe.

Overview

This recipe provides the following:

• Jobs that score and grade records for three Salesforce objects: Accounts, Contacts, and Leads.
• Bots to run the above jobs. *Note that the bot will be created but not automated.
• Manual data sources that provide metrics on what field values have been normalized.
• Analytics reports
• An App Factory Dashboard to view the analytics reports containing metrics of scores and grades.

Preparation

Establish data sources:  A data source must be created for all three Salesforce objects: Accounts, Contacts, Leads, and Campaign Members. You do not need to create new data sources if you already have these data sources, though they may need some editing to include the correct attributes.

Configure data sources: The data sources for Accounts, Contacts, Leads, and Campaign Members need to contain the following attributes:

*Note: All Blue Attributes relate to customer fields that will need to be created in Salesforce to hold the specific values and may have been created as part of the required Data Essentials, and Segmentation solutions that precede Scoring Implementation.
All Orange Attributes relate to fields that may or may not be included in your source system and can be omitted if they will not be needed in the overall scoring model.

1. Create new data targets: Create a new data target for each object with the following attributes. Note: If you don’t want to update Salesforce with all of these fields, you can customize the export task after the project has been created to update the specific fields desired.
   

   Account	Contact	Lead
   Account Grade	Contact Grade	Lead Grade
   Account Demographic Score	Contact Demographic Score	Lead Demographic Score
   Account Behavior Score	Contact Behavior Score	Lead Behavior Score

2. Configure reference tables: The solutions are linked to Openprise's Data Catalog, where sample scoring and grading tables are included. You can either download the available tables as a starting point, upload them to GoogleSheets or other supported cloud storage providers like Sharepoint, and then customize them to your company or start from scratch with a table in GoogleSheets or other supported cloud storage providers like Sharepoint, and then customize them to your company. If you choose to user your own scoring and grading tables, consider if you want to add an import step for them in the bots so the jobs are always using the most up to date tables.
   
   Your reference table should include the different criteria required to be assigned a specific grade or score. These will generally follow the structure listed below but can be made to handle more complex conditions: 
   

   Score or Grade	Field 1	Field 2	Field n
   Grade/Score 1	Condition 1	Condition 1	Condition 1
   Grade/Score 1	Condition 2	Condition 2	Condition 2
   Grade/Score 2	Condition 3	Condition 3	Condition 3

   *Note: If you are having difficulty configuring or scoping the conditions that should be included in your scoring table, please contact your CSM, who can assist with common strategies and complex condition creation. 

Clone

To clone the project recipe, navigate to Projects Recipes and select the project recipe to clone.

You will need to provide:

• Project name
• Prefix and/or suffix for all objects in the project
• Data sources to be used in the project *Created in preparation step above
• Data targets to be used in the project *Created in preparation step above

Examine the cloned project

Look through all the newly created project elements to familiarize yourself with what has been created.

Look for:

1. Any job and/or task that is marked with errors. These errors must be manually corrected.
2. Any tasks marked with [EDIT] in the task name. The description guide you on what to change for your environment.
3. Any tasks that write data back to Salesforce.  Make changes by unmapping any values that you do not want to write back to Salesforce.
   *Note: Any Export task to Salesforce has been bypassed for your protection. We recommend that you run the jobs with the bypass enabled to review any data changes before writing data back to Salesforce.

Run the bot

The first time you run the bots, we strongly suggest you keep the bypass option enabled on all export tasks or even remove the update jobs from the bots, treating Openprise like a sandbox test environment, until the output data looks correct.

Look at the output from each job, carefully examining your data to ensure it is being transformed correctly and to your standards. Try to think of some edge cases in the data that might pop up to ensure they are being transformed appropriately. Review each export task to verify the mapped attributes are set up correctly.

Test Exports to Salesforce Production Environment

When you are happy with the job outputs, complete two test runs by following the steps below.

• The first test should be on one record, and this will validate that Openprise has all the needed write permissions to update the Scores and Grades.
• The second test should be about 400 records. This test will validate if any APEX or Row_Lock errors might be encountered on the object, which might require a reduced batch size. Read more about setting custom batch sizes in the export task here. 

 Testing steps

1. Filtering the input of the export job, "SFDC Account, Contact, or Lead Scoring - 04 Update Lead Scores," to the test record(s) using the ID attribute for one record, and email begins with a batch of records.
2. Remove the Bypass option on the export task.
3. Remove the filter on the Log to MDS tasks.
4. Run the job.
5. When the job is complete, review the records updated by looking at the job outputs and reviewing the attributes op_export_success and op_export_errors. Please also review the records in Salesforce to verify the changes were made correctly.

Finalize the project

The final steps are:

1. Edit the export task to remove the restriction in the filter records task for the number of records to export.
2. Purge the jobs to prepare for a clean, automated run.
3. Edit the Bot to schedule the frequency and enable automation.
4. Add a bot to import your reference tables on a standard cadence to ensure your scoring and grading criteria in Openprise are always up-to-date with any edits made to the source tables in the Cloud Storage Provider.