You are looking at some records and wondering why were they are scored A, B, C, D or E grade while you'd expect another grade?

Try this quick checklist to see if it helps you explain the Lead Grade.

This article covers the Salesforce integration but would be similar with other integrations.

Auditing the Lead Grade

Breakdown of what goes into the Lead Grade and where to find the information

1. Do the Customer Fit and Likelihood to Buy scores match the Lead grade matrix?

• The Lead Grade is the result of the combination of the Customer Fit and the Likelihood to Buy model (LTB). Is there a problem with the Lead Grade here? 
  • How to: check the Lead grade matrix in your account. Go to app.madkudu.com > Predictions> on the Lead Grade section click "View performance" then "Diagnostics" button at the top right.  
  • Example: the record is scored with a Lead grade D from any of the 3 possible combinations below. Is it not the case?
    

• • If the Lead Grade doesn't match the matrix of Fit and Likelihood to Buy, please open a ticket here with 
    • examples of object/record ids
    • the grade and scores you see 
  • If the Lead grade matches, proceed to the next question. 

2. If expect that the Likelihood to Buy score should be different, what events has the lead performed? 

• Is it that the lead performed events but don't seem to be included in the score? 
• Check the Likelihood to Buy signals for the lead.
• Are these events included in the event mapping?
  • The event mapping is the backbone of the Likelihood to Buy model and controls all the events that are ingested in the Likelihood to Buy model.  
  • How to:
    • check the Event mapping in the App: section Mapping > Event mapping. Select the connector where your event is tracked, and check if it is mapped into an event in MadKudu platform. 
    • If it is not present, this means this event is not included in the scoring. You will need to add this event to the event mapping and update the model to assign points to this event. 
    • If it is present, proceed to the next question.
• Are these events included in the model?
  • An event may be present in the event mapping but not necessarily associated with points in the model.
  • How to: check the Likelihood to Buy model in the Data Studio, the section "Event weights": do you see the event in the model?
    • If no, it could mean
      • the event was added to the mapping but the model was not updated to have the event appear in the Data Studio. 
      • the logic in the event mapping isn't correct.
        Either way, you would need to update the Likelihood to Buy model with this event. 
    • If yes, proceed to next question.
• Are these events scored high enough to give a high score?
  • If you expect leads performing these actions to be scored higher, check their weight and decay. Understand how the Likelihood to Buy score is calculated. 
  • How to:
    • in Data Studio's section "Event weights", check the score of the event. 
    • in Data Studio's section "Thresholds", check the thresholds to become a medium, high, very high Likelihood to Buy 
    • Change the weight and decay according to your need with these instructions. 

3. If you expect a different Customer Fit score, what is the enrichment associated with this record? 

• You can check which enrichment data points go into your Customer Fit model  or check out how the score is calculated by looking at this article. 

4. Was the record recently updated by MadKudu? 

• If you are expecting a different score for this record than you actually see, it may be because the record was not (yet) updated.
• How to:
  • check the record update history in Salesforce
  • please refer to this article to troubleshooting missing or delayed updates

5. Haven't figured it out or think there is more to it? 

Please open a ticket here with the following information:

• a few records ID 
• what you see (score, signals, events, enrichment ...)
• what you expect to be different for these records and why
• anything else you've already looked into 

so we can help you out in a timely manner.