Do the scores work by default?
Yes, the scores are calculated 24 hours after your synced accounts are enabled and the user clicks the Save and train model button on the Predictive Score Setup page. At least one Keyword Set needs to be created.
To access this page, from the left navigation bar, go to Settings > Demandbase Wide-Settings > Predictive Score Setup.
How long does it take the model to run?
24 hours
What is a good score?
75+ is worthy of specific attention and 95+ requires immediate attention.
Can you revert to an earlier model that was run for the Pipeline Predict and Qualification Scores?
No, you need to be sure before retraining the model as there's no mechanism for reverting to an earlier version.
How often does Demandbase recommend that I retrain the Predictive Score Models?
We recommend you enable the Auto Retrain option to make sure your predictive models are always up-to-date.
If you prefer to update the scores manually, retrain Pipeline Predict Scores every 3 months and Qualification Scores every 6 months.
How do I see what factors influence the Predictive Score for a specific account?
Go to the Analytics dashboard and scroll down to the Accounts section. On the Pipeline Predict or Qualification Score tab, hover over the score labels (e.g. All, Advertising) of the account you want to view. A window appears that shows the top factors determining the score for this account.
How does Predictive Score improve over time?
The models automatically pick up new signals every night from all available sources.
Are Pipeline Predict Scores dynamic? How frequently do the models change/pick up account activity?
Pipeline Predict is updated for all accounts nightly as new activities appear for each account. The Pipeline Predict Score can also decrease when existing activities become stale.
Does the Pipeline Predict Score only look at engagement from the Marketing Automation System (MAS)?...
...What about other engagement types that the MAS is able to capture? I recognize that some MAS are more robust than others. So perhaps email activity represents the lowest common denominator. But for those that have richer engagement details available (e.g. Marketo), does Demandbase use those signals as well?
Anything that is an activity that you can assign Engagement Minutes in the platform are used by the Pipeline Predict Score.
How does the Pipeline Predict Score know which Personas/Titles are more relevant?
The model decides automatically based on all the data plus the “level” of the title (CMO → VP → Director → Manager).
What’s the difference between High, Medium, Low Intent?
Demandbase Intent measures the strength of intent by the number of people at similar size accounts reading web pages with a keyword.
Important: This methodology and section may be updated as the team makes improvements to the model.
| Account Size | Low | Medium | High |
|---|---|---|---|
|
1 - 10 employees |
1 person reading | 2 - 3 people reading | 4 or more people reading |
|
11 - 50 employees |
1 person | 2 - 4 people | 5 or more people |
|
51 - 200 employees |
1 - 2 people | 3 - 7 people | 8 or more people |
|
201 - 500 employees |
1 - 2 people | 3 -9 people | 10 or more people |
|
501 - 1K employees |
1 -2 people | 3 - 13 people | 14 or more people |
|
1K - 5K employees |
1 - 3 people | 4 - 18 people | 19 or more people |
|
5K - 10K employees |
1 - 3 people | 4 - 28 people | 29 or more people |
|
> 10K employees |
1 - 3 people | 4 - 37 people | 38 or more people |
If my company integrates with CRM but wants to add a new list of accounts via a CSV file, will they get a consistent score across both the CRM- and CSV-generated accounts?
All accounts are scored the same way, regardless of where they came from.
If my company is uploading a CSV with opportunity data, what is the minimum level of detail required to get a good model in place?
The model only needs Account Website/Domain and Created Date. There are additional required fields for Data Import, but those fields are used by other features in the platform.