Pipeline Predict Scores indicates how likely an account is to become a pipeline opportunity in the near future. This metric helps identify accounts that are ready for targeting, supports pipeline planning, and signals when it's the right time for sales outreach.
Additionally, the score can show when existing customers are likely to consider growth (renewal, upsell, or cross-sell) opportunities in the near future. This provides actionable insights for growth strategies, ensuring you maximize opportunities within your customer base.
See Set Up Pipeline Predict Scores to configure your Pipeline Predict Scores.
Tip: You can also set up multiple Pipeline Predictive Scores for your different product offerings. See Understanding Multiple Predictive Scoresfor more information.
How Does It Work?
Pipeline Predict is a machine learning model that looks at past opportunities, learns account properties and activity patterns that are leading indicators of new opportunities being created at accounts, and then identifies those leading indicators in other accounts.
Quick Facts
Score |
|
Initial Model Build Time | 24 hours |
Model Re-Training Time |
24 hours |
Score Update | Automatic (nightly) |
Customer Connection Modes |
|
Factors Used When Calculating Pipeline Predict Score
Demandbase uses the following data to calculate Pipeline Predict Scores:
Past Opportunities |
New business opportunities opened within the past 12 months, from at least 50 distinct accounts. See Opportunity Criteria for Pipeline Predict for in-depth information. Tip: Out-of-the-box settings use opportunity Created Date in the past 12 months. |
All Activity Logged in CRM (account/contacts/leads) |
Uses activity type/category and role. It doesn't use activity details. |
Engagement Minutes |
Uses activity type/category and role. It doesn't use activity details. Important: Pipeline Predict Score only uses activities with Engagement Minutes greater than 0. |
Sales Inbox/Calendar Events |
Uses email open and click and calendar events. It doesn’t use the title or details. |
MAS: All email activity |
Uses email open and click. It doesn’t use the title or content. |
Firmographics | Uses country, state, industry, revenue range, and number of employees. |
Trending Intent (daily update) |
Uses daily intent signals. |
Historical Intent |
Uses intent behavior up to one year prior to the opportunity Created Date. |
Known Website Visitors | Uses which job role visits the website and their number of visits. It doesn’t use what pages were visited. |
Unknown Website Visitors |
Uses the number of unknown page visits. It doesn't use what pages were visited. |
Example
The following image shows an example of how the Pipeline Predict Score is calculated for an account.
Opportunity Criteria for Pipeline Predict Model
The Pipeline Predict model requires opportunities that meet the following criteria:
- At least 50 distinct accounts that opened New Business opportunities. See New Business Opportunity Definition.
-
Opportunities opened within the past 12 months.
Tip: Date range selector can be set to less than 12 months. However, we recommend using a full 12-month window to ensure you’re capturing enough data for accurate analysis.
New Business Opportunity Definition
At the time the opportunity was created, the account was NOT in a “Customer” designated stage.
In cases where journey stages are not explicitly designated, the model infers the designation using keywords within the stage name.
Important: The model uses stage designation settings from the Global journey funnel
An account is considered to be in a “Customer” stage if either of the following conditions are true:
-
The journey stage is designated as “Customer”.
OR -
The journey stage name:
-
Contains the keywords customer, expansion, closed/won, existing, upsell, renewal, client, or churn risk.
AND - Does not contain the keywords lost, not customer, ideal customer profile, or ideal client profile.
-
Contains the keywords customer, expansion, closed/won, existing, upsell, renewal, client, or churn risk.
Accessing Pipeline Predict Score Results
After an Administrator configures the Pipeline Predict Score, you can access Pipeline Predict score results. See Understanding the Account List Dashboard.
Retrain Pipeline Predict Scores
Scores must be retrained at least once a year to make sure the model uses the most recent historical data. You can retrain scores using the Auto or Manual Retrain option. To increase score accuracy, we highly recommend using Auto Retrain.
Important: The Auto Retrain option is enabled by default.
-
Auto Retrain (recommended): Demandbase automatically retrains the Pipeline Predict Score every 3 months. If you edit and retrain a model, the auto retrain schedule starts on the date you retrained the model.
Tip: To auto retrain Pipeline Predict Scores, from the left navigation bar, goSettings > Predictive Score Setup. Click the name of a Pipeline Predict Score and click the Configure tab. Toggle on Auto Retrain to enable the auto retrain feature.
-
Manual Retrain: Admins manually retrain the Pipeline Predict Score. We recommend retraining Pipeline Predict Scores every 3 months.
Tip: To manually retrain Pipeline Predict Scores, from the left navigation bar, goSettings > Predictive Score Setup. Click the name of a Pipeline Predict Score and click the Configure tab. Make edits and click Save and train model.