Do you have a fit model that scores your leads and contacts, and would you like to apply a similar logic to score accounts? Read the following article to understand the different ways to create an account fit model.

What is an Account Fit?

The Account Fit model scores accounts based on the DNA of the company (firmographic and technographic data). MadKudu uses the domain of the account to perform identity resolution.

The model learns from your past conversions to identify which accounts in your database look like your current or past customers.

How do I create an Account Fit?

MadKudu Customer Fit models created in your Data Studio can score leads, contacts, or accounts. The same scoring mechanism is used for all these objects; the only difference between an Account model and a Person model is the object to which the scoring logic is applied.

Option #1 - You want your person-level customer fit to also score your accounts

MadKudu can easily push your person-level customer fit model to your accounts. This model can integrate many different attributes, such as technographic, firmographic, and demographic.

Before pushing your model to your account, ensure it is not overly reliant on attributes only present at the person level, such as title, person country, and seniority. Otherwise, the account level scoring will be incomplete, based only on the attributes related to the company.

Review your live customer fit model in the Data Studio and check whether person-level attributes are an important part of the scoring logic (i.e. those attributes are used in many parts of your model trees, overrides, etc.). You should look for:

• In trees, person-level computations that are placed high in the nodes.
• Overrides boosting leads based on firmographic computations.
• Account-level signals that are related to a person. For example, "Lead is using a personal email" can be changed to "Domain is personal".

Create the account fit score fields in your Salesforce.

Then, open a support request here so that our team can finish the custom set up and start pushing the account fit.

Option #2 - You want to create your specific account-level model

This option can be chosen when you think your scoring logic for accounts should differ from that of leads and contacts.

Apply the same mechanics as when creating a person-level customer fit model. Remember to not base your model on any lead-level attributes such as title, country, and seniority.

Useful guides to create your account fit from scratch📚

• Customer Fit: Training and Validation datasets
• Customer Fit: How to Create or Edit a Decision Tree
• Customer Fit: Signals
• Customer Fit: Overrides

Once you have created your account fit model, please open a support request here so that we can start pushing it to your integrations.

What's next?

After creating your account model, either through option #1 or #2, open a support request to begin pushing the account fit to your integrations.