MadKudu helps you remove barriers between your siloed sources of data help you segment and target your highest-value prospects.
How does the scoring work? What data points are used and when? How does MadKudu reconcile different enrichment points to one person or one company? We hope this article will answer these questions.
Lead and Contact Fit scoring
MadKudu turns an email into a full person and company profile to predict if this person is a good fit for your product.
Real time scoring
- Any new Lead or Contact is scored in real-time by our API 10-15min after they are created in Salesforce by your workflow.
- Leads and Contacts are scored in real-time based on
- the standard Salesforce field email
- the enrichment MadKudu has on this email from our own sources. Learn more
- any enrichment you may already have in your Salesforce about this domain (MadKudu is able to cache information from your Salesforce accounts in real time based on their domain).
If the email field is empty, MadKudu won't be able to score this record.
- Then in batch, MadKudu updates the score with additional enrichment available on the Lead and Contact record in your Salesforce. (MadKudu does not have access to this information in real time).
- email@example.com downloads some content on your website and has input in the gated form (CPO, team size = 10). You don't know Francis nor MadKudu yet in your Salesforce. Your workflow now creates a Lead for Francis in Salesforce.
- MadKudu scores this Lead "75, good" within 15 min after its creation, based only on the enrichment MadKudu has from our vendors.
- Within 4-12hours, MadKudu will update the score of the Lead with additional information on his title and his team size (at the next Sync process)
- MadKudu now scores this "Lead 85, very good".
- If there is any change from the MadKudu enrichment or your CRM enrichment, then the score would reflect this change within 4-12hours (at the next Sync process) from when it is used in the model.
How does MadKudu use CRM Account attributes to score Emails?
Account Fit scoring
MadKudu turns a domain into a full person and company profile to predict if this company is a good fit for your product.
- When the Account is created, MadKudu scores the record in batch within 4-8hours of its creation based on
- the standard Salesforce field "website".
- the enrichment MadKudu has on this domain from our own sources.
- with enrichment available on this Account in your Salesforce.
- MadKudu does not score in real time accounts, only in Batch.
- If this field is empty, MadKudu won't be able to score this record and you would see a null score.
- You have added to Salesforce the Account acme.com with no other data point.
- MadKudu scores this Account "81, good" within 4-8h after its creation, based on the enrichment MadKudu has from our vendors and the information present on the Account record (in this case, none).
- Your team now marks this account as "target account" by flagging it through your custom field is_target_account__c = 'true'
- MadKudu will update the score of the Account within 4-8h, let's say now to "95, very good" because this field is_target_account__c is used in the model to boost target account.
- If there is any change from the MadKudu enrichment or your CRM enrichment, then the score would reflect this change within 4-8hours from when it is used in the model.
MadKudu only uses certain data points, not all of your CRM.
MadKudu only uses certain data points from our sources and your CRM to score the Lead or Contact. This selection is done when building the model when our team looks at which data points would be relevant and would have a value to include in the model. We can add more data points in the model or as overrides. For that, we would assess the need and possible incremental value to add these data points.
How to know which attributes from my CRM are used in a model?
To see the list of data points used in your Customer Fit model, please refer to:
the Attribute mapping in app.madkudu.com > Mapping > Attribute mapping will tell you which fields are specifically mapped from your CRM and potentially used in the model*. Learn more about Attribute mapping.
*Note: not all fields mapped in the Attribute mapping are used directly in the model. They may have been mapped to understand their impact on conversion in the data science studio but not necessarily relevant to use in the scoring model.
the Computations page in the Data Science Studio https://springbok.madkudu.com/ > Customer Profile > Computations has computations marked as "LIVE" meaning they are used in the customer fit scoring. Learn more about Computations.
or get in contact with our support team at firstname.lastname@example.org to help you figure that out.
How often are the scores updated?
- Within 15min of the Lead or Contact creation and then updated 4-12 hours (at the next Sync process) later with additional data points from your CRM. Please refer to this article: https://support.madkudu.com/hc/en-us/articles/360036910411-How-long-does-it-take-to-score-a-lead-
- For any change of the enrichment data points used in your model
[Salesforce] What happens when a lead is converted into a contact?
- When a lead gets converted into a contact, MadKudu will automatically score the contact with the same attributes as the lead + any new attribute on the contact. When a SFDC Lead is converted to a Contact, the originating Lead object is locked and no further updates are permitted - this is by SFDC design.
What data does MadKudu use if the Lead and Contact have different enrichment?
- MadKudu would gather all possible data points from all records with the same email and take the MAX(field). This allows us to capture all non empty values. For edge cases where the enrichment differs for non numeric fields, the limitation of this methodology is that it takes the latest value in alphabetic order. We understand this is not ideal but recommend to resolve discrepancies by updating to the correct enrichment or deleting the incorrect enrichment.
What data does MadKudu use if I have 2 different Accounts with different enrichment?
- The customer fit model scores people and companies based on who they are, their "DNA" to understand if they are intrinsically a good fit for your product. Therefore differences of enrichment for 1 same email or 1 same domain in your CRM would most likely come from duplicated records and we recommend resolving these duplicates or enrichment discrepancies to make sure the score would be more accurate.
- When there are 2 accounts with different enrichment we take the MAX(field) (please see the answer right above)
How does MadKudu select which data to use in priority when there are multiple sources of enrichment?
- By default, we use the MadKudu enrichment in priority and if empty we then use data from your CRM. Some companies have less trust in their own enrichment than the MadKudu enrichment when data is manually inputted by their team and prefer this prioritization. If you would prefer to prioritize the enrichment of your CRM over MadKudu's enrichment, this is something that can be changed to fit your need and we'll be happy to assist. Please reach out to email@example.com for assistance.