To make things easier for the end-users of the scoring (your Marketing and Sales teams), you may want to use a label (or "segment") instead of a score. For example, does your team know if a score of 20 is a good score because it's 20 out of 30 or it's a bad score because it's 20 out of 100?
To set the definition of what is a very good, good, medium or low score, you can do it in the Data Science Studio.
MadKudu fields in Salesforce
Labeling thresholds configuration in the Data Science Studio
How to configure the score thresholds for the Customer Fit Segment?
- You have the permissions of the Architect role
- You are working on a point-based model
Step 1: Duplicate the live model
- Go to the Data Science Studio (springbok.madkudu.com)
- Duplicate the model marked as "live" so you don't risk editing the model currently in production
- Name the duplicated model how you want
Step 2: Change the thresholds
- Click on Data Science Studio at the top of the page
- Click on the Ensembling tab
The Ensembling tab allows you to configure the 4 available segments: very good, good, medium, low. You can set the thresholds defining what is the minimum score for each threshold.
For example here, any lead or account with a score
- above 150 (included) will have a customer fit segment Very good
- above 80 (included) but strictly below 150 (i.e. between 80 and 149) will have a customer fit segment Good
- above 30 (included) but strictly below 80 (i.e. between 30 and 79) will have a customer fit segment Medium
- below 30 will have a customer fit segment Low
- Set the thresholds and click Compute
- Iterate on the thresholds to make sure you have a distribution of leads or accounts looking like
- 10% of very good
- 20% of good
- 30% of medium
- 40% of low
so that your team only focuses on 10-30% of qualified accounts
Step 3: Deploy changes
To make the changes active in your scoring and start updating the segment of your leads or accounts, you'll need to deploy the changes made.
- Coming soon for Architects! In the meantime, please send a request to email@example.com