Create Pipeline Predict Scores

  • Updated

Pipeline Predict Scores indicate how likely an account is to become a pipeline opportunity in the near future. You can create scores for new business opportunities or growth opportunities with existing customers.

Important:

  • Changes to Pipeline Predict Scores can have platform-wide implications. Model updates can take up to 24 hours. Changes take effect after the model is retrained and account scores are updated. You cannot make additional changes until the update is complete.

  • Retrain scores whenever you update your engagement points or intent keyword configurations. Auto Retrain is enabled by default. See Retrain Pipeline Predict Scores.


Opportunity Types for Pipeline Predict Scores

You can create Pipeline Predict scores for the following opportunity types:

Opportunity Type Definition Recommendations
New Business Opportunities with new customers. Use opportunities opened in the past 12 months. Expand the date range if you have longer sales cycles.
Growth Renewal, upsell, or cross-sell opportunities with existing customers.

Use opportunities opened in the past 12 months. Expand the date range if you have longer sales cycles.

Include past and current customer accounts. For example, use the selector Account Type = Customer.


Prerequisites to Create Pipeline Predict Scores 


Create a Pipeline Predict Score

  1. From the left navigation bar, go to Settings_N.png Settings > Demandbase-Wide Settings > Predictive Score Setup and click Create New.
  2. Enter a descriptive score name, select Pipeline Predict, and click Next.
    Tip: Use descriptive names to differentiate scores. For example, Product A New Business, Product A Growth, or Product B New Business.
    New_Pipeline_Predict.png

  3. In the Past Opportunities section, add selectors to identify opportunities to use for training the model. The selected opportunities must meet the following requirements:
    • Include opportunities from at least 50 distinct accounts.
    • Include opportunities opened within the past 12 months, or a longer period for longer sales cycles.
    • Can include open, closed-won, and closed-lost opportunities.
  4. In the Engagement activities section, add selectors to identify the engagement activities to use for training the model. For example, select activities related to a specific product.
    Important: Do not use Activity Date filters in the Engagement activities selector. See Best Practices for Setting Up and Using Predictive Scores.
  5. (Optional) Toggle on Auto Retrain to automatically retrain the model every three months.
    Tip: We recommend keeping Auto Retrain enabled. See Retrain Pipeline Predict Score.
  6. Click Save and train model.
    Tip: Before saving, click Refresh to see how many accounts meet your selector criteria.configurePP.png

Delete Pipeline Predict Scores

  1. From the left navigation bar, go to Settings_N.png Settings > Demandbase-Wide Settings > Predictive Score Setup.
  2. Select the checkboxes next to the Pipeline Predict Scores you want to delete.
    Important: You cannot delete the All Pipeline Predict Score.

  3. Click Take Action.
  4. Click Delete.
  5. Click Delete Score to confirm.


View Pipeline Predict Score Results

After the model finishes training, you can review Pipeline Predict Scores to identify accounts most likely to become opportunities.

  1. From the left navigation bar, go to Analytics > Dashboard.
  2. Select an account list from the drop-down.
  3. Scroll to the Accounts section and click the Pipeline Predict tab.

Accounts labeled as Highly Likely have high engagement and are likely to become opportunities. Prioritize these accounts for sales outreach.

See Take Action for ways to engage with recommended accounts.

PP Next Steps.png