Set Up Pipeline Predict Scores

Context

Pipeline Predict is a machine learning model that looks at past opportunities from your CRM, learns account properties and activity patterns that are leading indicators of new CRM opportunities being created at accounts, and then identifies those leading indicators in other accounts. The score indicates how likely it is that a new CRM opportunity will be created at the account.

Pipeline Predict takes the following into consideration:

• Past opportunities from CRM
• All available data within CDP
  • Email Activity from MAS including opens/clicks/etc.
  • All activity logged in CRM for each person in the account
  • Sales Inbox Activity
• Firmographics
• Intent Data, Historical and Trending
• Website Visits
• Advertising Activity

Important: Changes to the model can have major implications platform-wide. The model takes up to 24 hours to update. Any changes are reflected after the model has been re-trained and each account score is updated. You cannot make any additional changes until the update process is complete.

Preparation/Prerequisites

Review the following articles:

• Understanding Pipeline Predict Scores
• Working with Selectors
• Understanding Activities, Intent, Engagement, and Data

You must have Admin privileges. 

Pipeline Predict scores require either a CRM integration or an upload of open opportunities via a CSV file.

To upload to open opportunities, Pipeline Predict scores leverage the Create Data and (<13mo) and Actual Website fields. The Owner field is useful for filtering opportunities.

Steps

1. From the left navigation bar, go to Settings > Demandbase-Wide Settings > Predictive Score Setup and click Create New.
2. Enter an appropriate Predictive score name (for example, the name of the product that the score is for), select Pipeline Predict, and click Next.
   
   The Pipeline Predict setup page opens.
   
   
3. In the Past Opportunities section, add Selectors to filter for the past pipeline opportunities to use to train the new Pipeline Predict Score. 
4. In the Engagement activities section, add Selectors to filter for engagement activities that the Pipeline Predict Score can use as scoring criteria.
5. Click Save and train model.
   Tip: Before saving the Pipeline Predict Score, click Refresh to see the number of accounts that meet the Selector criteria you entered.

Outcome

The model takes up to 24 hours to update.

Next Steps

Access the Pipeline Predict score results and take action on the recommended accounts.

A Pipeline Predict score between 50 and 90 indicates an account is likely to become a CRM opportunity. A score above 90 indicates an account is highly likely to become a CRM opportunity.

To access the score results, from the left navigation bar, go to Analytics > Dashboard and select an account list from the drop-down list. Scroll down to the Accounts section and click the Pipeline Predict tab.