Account Scoring models

Algorithm name: Account Scoring

Applies to: B2B

Account scoring algorithm helps identify accounts/ opportunities at different stages of the sale funnel and evaluate their likelihood of making a product purchase/ subscription. This helps marketers identify conversation-ready opportunities.

Also commonly known as: Opportunity scoring, Opportunity qualification

What is Account scoring algorithm?

The Account scoring algorithm scores B2B accounts on their likelihood to purchase or their likelihood to subscribe to a product. The model uses a combination of firmographic attributes, contacts’ information and their engagement behavior to assign a numerical score to each account. Account scores are timestamped for every account, allowing you to more effectively target account segments and align sales and marketing strategies.

Parameters of the model

To create and configure the model the following parameters must be defined:

  • Algorithm: Choose the Predictive Account Scoring B2B algorithm in Unity.

    • Lookback window: You must also set a lookback window (number of days to analyze historical data)

  • Queries: The queries you select generate the dataset used for both model training and scoring.

  • Inputs: Inputs are attributes from the Oracle Unity data model that the model uses during training and scoring.

  • Outputs: These are the data objects and attributes from the Unity data model that store the model's output values. You can customize the default output mappings if needed.

Model inputs

The Account scoring model uses the following data. For the model to successfully run, check the sections: key input considerations, key data guidelines and best practices.

Attribute from Unity data object Attribute name in the query Unity Data object Description of the attribute Data Type Must have?
AnnualRevenue annualrevenue Account/ MasterAccount Annual Revenue of the company Float  
EmployeeTotal employeetotal Account/ MasterAccount No. of employees in the company Int  
LineOfBusiness lineofbusiness Account/ MasterAccount Industry of the contact String  
NAICSCode naicscode Account/ MasterAccount North American Industry Classification System Code String  
SICCode siccode Account/ MasterAccount Standard Industrial Classification Code String  
SourceAccountID/ SourceMasterAccountID sourceaccountid Account/ MasterAccount Unique identifier for the account String Yes
SourceID sourceid Account Source ID of the Account String Yes
YearEstablished yearestablished Account/ MasterAccount Year of establishment of the company String  
Email email Customer Contact's Email Domain String  
JobTitle jobtitle Customer Title of the Contact. e.g. VP, Manager, Analyst etc. String  
JobTitleLevel jobtitlelevel Customer Level assigned to each job title String  
okToCall oktocall Customer Flag reflecting whether the contact is Okay to be contacted via Call String  
okToEmail oktomail Customer Flag reflecting whether the contact is Okay to be contacted via mail String  
okToMail oktoemail Customer Flag reflecting whether the contact is Okay to be contacted via Email String  
okToNotify oktonotify Customer Flag reflecting whether the contact is Okay to be contacted via Push Notification String  
okToText oktotext Customer Flag reflecting whether the contact is Okay to be contacted via Text String  
SourceCustomerID sourcecustomerid Customer Unique identifier for the contact String Yes
AvgCompositeScore avg_composite_score Event Intent score, if available Float  
EventTS eventts Event Timestamp of when an event occurred Timestamp Yes
Medium medium Event Medium through which the activity happened. e.g. email etc. String Yes
Type type Event Type of the event. e.g. Open, Sent, Clicked etc. String Yes
City city MasterAccount/ Address City of the Contact String  
Country country MasterAccount/ Address Country of the Contact String  
State state MasterAccount/ Address State of the Contact String  
ZipCode zipcode MasterAccount/ Address Zip Code where Contact resides String  
IS_CONVERTED/ SalesStage is_converted Opportunity 1 if there was a conversion, 0 otherwise Boolean/Int Yes
OpportunityDate opportunitydate Opportunity The date when a particular the opportunity converted Timestamp Yes
ProductID productid Opportunity ID of the Product String Yes
ProductName productname Opportunity Name of the Product String  

Key considerations on model inputs

  1. Data schema checks: If an attribute is unavailable, assign a constant default value across all records. This allows the model to validate the input schema. This will not impact the model outcomes as the attribute will be excluded during feature importance analysis due to lack of data variance.

  2. Product levels: If you have a product hierarchy maintained in the data, choose the right level to model at and map it to the ProductID attribute in the query.

  3. Schema consistency: Ensure that the attributes used in the query exactly match the specified 'Attribute name in the query' (in the above table) to avoid schema mismatches during model execution.

  4. Non-mandatory attributes: These attributes, when available, can enhance the model’s learning and improve its predictive performance. However, if they are missing, the model will still leverage other available attributes to make predictions.

Key input data guidelines

  1. Label requirements: Ensure the model data contains at least 500 positive signals (converted accounts) and 500 negative signals (non-converted accounts)

  2. Data requirements:

    1. It is recommended to provide at least 50,000 data records for model training. However, it is important to bring as much data as possible to build a robust model

    2. It is recommended to keep the dataset under 10M records (soft limit). This limit can vary for every model, but maintain a manageable size helps ensure efficient model performance.

Best practices

  1. Use case first approach: Begin with a specific use case to determine the data required for solving the problem effectively.

    1. Review model data requirements with both business and data teams to ensure alignment

  2. Contextual parameters: Choose model parameters (such as lookback window) based on the business/ use case context.

  3. Query validation: Test both the training and scoring queries to validate that they return data and inspect sample records resulting from the query

Model outputs

Output values will be stored in the AccountProductScore data object. You can review the account score values for each account in the Status attribute. Each status (cold, cool, warm, and hot) is based on the account score generated from 0 to 100.

Attribute ID Attribute Name Attribute Description   Data type
SourceAccountProductScoreID Source AccountProductScore ID This attribute contains the unique ID for the table. STRING  
ScoreTS Score Timestamp This attribute represents time when the score was computed. TIMESTAMP  
Score Score This attribute represents score for the account product. INT  
Status Status This attribute represents account status - Cold, Cool, Warm, Hot. STRING  
ProductID Product ID This attribute contains the foreign key to the ProductID. STRING  
SourceProductID Source Product ID This attribute contains the original form of the Product ID from the source data system. STRING  
ProductName Product Name This attribute contains the product Name. STRING  
ProductIDoutput ProductID Output level This attribute contains the product ID (if the output level is different). STRING  
MasterAccountID Master Account ID This attribute contains the foreign key to the MasterAccountID. STRING  
SourceMasterAccountID Source Master Account ID This attribute contains the original form of the Master Account ID from the source data system. STRING  

The Status attribute will generate one of the following values based on the account score.

Account score Status Sample interpretation
0–25 Cold Low engagement/ intent; deprioritize
26–50 Cool Place in long-term nurture segment
51-75 Warm Enrol in high-touch nurture campaigns
76-100 Hot Route to Sales

data science, data science model, analyze data, create data science model, how to create a data science model, predictive contact lead scoring, predictive contact lead scoring model, lead scoring model, lead score, lead scoring, b2b model, b2b lead scoring, b2b lead scoring model