Overview of the method
- You have an Architect or Admin MadKudu role
- You know what an Override is
Summary of the steps
MadKudu processes your data in a sequence of steps to score your records in your CRM.
To add an override to a customer fit model which uses a field from your CRM, you will need to configure several steps of this data pipeline.
- Step 1: Pull fields from your CRM into MadKudu platform
- Step 2: Map the fields in the Attribute mapping
- Step 3: Create a computation from the Attribute
- Step 4: Create the Override
- Step 5: Check the performance of the model
- Step 6: Update signals (Optional)
- Step 7: Review
- Step 8: Deploy the model and announce it to the Sales team
Step by Step
Step 1: Pull field(s) from your CRM into MadKudu
app.madkudu.com > Integrations > Salesforce > Pull > Configure
- Pull from your Salesforce into MadKudu relevant
Accountfields to use in the overrides following these instructions.
- You'll receive a confirmation email when the data is pulled and ready to be used.
Step 2: Map the fields in the Attribute mapping
app.madkudu.com > Mapping > Attribute mapping
- In your MadKudu app account, map the Salesforce fields containing the relevant enrichment points in the Attribute mapping to make them available in the Data Studio.
- You'll receive a confirmation email when the mapping is executed.
Step 3: Create a computation for each of the fields
- Open Data Studio (or go to Predictions > Data Studio)
- Go to Computations to create a Computation
- Type: depends on your field type
- Name: field_lead_salesforce
- Configuration mode: OneToOne
- Field: select the attribute you created in step 1
Tip: If you have an field Industry on both the Lead, Contact and the Account, use the Configuration mode "Advanced mode" and write
COALESCE (attribute1,attribute2, attribute3...)
COALESCE (salesforce_lead_industry, salesforce_contact_industry, salesforce_account_industry)which means "get the value of the industry first in the Lead record, otherwise in the Contact record, otherwise the Account record it is attached too" -> the order is important.
- Repeat the same for each field you'll want to use in the override
- Once created, Click on Deploy
Step 4: Update the model dataset enrichment to include the new computation
Open the model you'd like to add the override to (or duplicate the live model).
In the Overview section, click on Reload computation to enrich the dataset with your new computation.
Step 5: Create the override
Then add the Override you need following these instructions
Step 6: Check the performance of the model
Before deploying your changes you'd like to make sure the override does not boost or penalize too many leads otherwise you may flood your marketing and sales teams with too many new leads or deprive them of good ones. Check the performance of the model in the Review > Performance to make sure the performance still looks good.
Open in a separate tab of your browser the model live you duplicated the model from and compare side by side whether the performance has greatly decreased in terms of conversion rates for each segment (refer to the second section in the screenshot).
(Optional) Step 7: Update signals
If you are adding an override you may want to show in the customer fit signals your Sales team see.
- Go to the Signals tab and add Signals based on the computations you've used in the Override
Step 8: Review
Before deploying any change to production, review a sample of leads in the Sample tab, see if you can find records impacted by the override.
Step 9: Deploy the model and announce it to the Sales team
Now that the override has been added to the model, you will need to push this update to your model in production to update the scores in Salesforce.
Keep these considerations in mind when deploying overrides:
- All your Leads, Contacts, or Accounts that are usually scored by MadKudu will get rescored with the next batch scoring within the next 4-12 hours (see when the next Analysis then Sync process should run in the Processes Page).
- Adding, editing, or deleting overrides that increase or decrease prospect scores may trigger automated workflows in your CRM which are based on the customer fit or lead grade score (like your MQL workflow).
- Don't forget to give a heads up to your Marketing and Sales team about the override you added!
When you are ready, go to the Deploy tab
- If the live model you are editing is flagged "Live as Standard", then go into the first sub-tab "deploy as standard" and click on deploy as standard fit
- If the live model you are editing is flagged "Live as Multi-fit" please open a ticket here, we'll help you with that.