Junk Score

This service helps you maintain high-quality data by evaluating and scoring potential junk entries in your dataset. By identifying invalid or low-quality data, such as incomplete, duplicate, or irrelevant records, it ensures that only accurate and useful information is retained.

Service Name: Junk Score

Service Type: Basic Utility Service

Service Description

You can use the Junk Score service to assess the likelihood of a profile being genuine or potentially problematic based on various profile attributes such as first name, last name, email, company name, and job title.  

The Junk Score service applies rules and algorithms based on various parameters:

  1. Data Validation: Checks for suspicious or irregular patterns in the provided information (e.g., random characters in the name fields, unusual email patterns).
  2. Comparison to Known Patterns: Compares the provided data with known patterns of junk profiles (e.g., commonly used disposable email domains, frequently used fake names).
  3.  Algorithmic Scoring: Utilizes an algorithm to assign weighted scores to different profile attributes based on their relevance and susceptibility to being associated with junk profiles.

The Junk Score service generates an overall score (junk value) based on the applied rules and algorithms.

Scores range from 0 to 100:

  1. Junk score 0 indicates a highly favorable, authentic profile.
  2. Junk score above 50 indicates a high likelihood of the profile being junk or problematic.
  3. Scores above 0 and below 50 signify a low/medium risk of the profile being potentially problematic or less genuine i.e. they can be reviewed

For example. there is a profile with the below details  

  1. First Name with the value “Geoff”, 
  2. Last Name with the value “Schmidt”, 
  3. Email with the value “gschmidt@directdigitalholdings.com”, 
  4. Company Name with the value “Direct Digital Holdings” and
  5. Job Title with the value “Partner and Senior Software Engineer”. 

When the above-mentioned profile will be passed through the Junk Score service, the service will return below output.

				
					Output- "result": {
		"score": 15,
		"rulesApplied": {
			"LNAME-CONSECUTIVE-CONSONANTS-VOWELS": "LastName contain consecutive consonants or vowels|(schm): 15"
		}
	}

Here the rule applied is LNAME-CONSECUTIVE-CONSONANTS-VOWELS: LastName contain consecutive consonants or vowels|(schm): 15 which has assigned score value 15.

Please note that the request parameters values (in step 2) and Marketo Fields (in step 4) are just for example. You have to enter the values as per your service requirement.

Steps to configure service Junk Score in Marketo

1. Go to Admin and click Webhooks. Click New Webhook.

2. Name and configure your webhook for service Junk Score.

Webhook Name: Name of the webhook. In the above example, the Webhook name is the same as the Service name. Webhook name can be provided as per requirement.

URL: The URL of the service Junk Score. The URL has domain and query parameters which will remain the same. You just have to change the parameter values with Marketo Token. To insert a token, click Insert Token.

Template: It will remain blank.

Request Token Encoding: For the POST method it will remain None, and for GET this field will be disabled and the default value of the field will be Form/Url.

Response type: Select the response format as JSON.

Request Type: The Request Type will remain the same (i.e GET) as given in the example.

3. Add Authentication Header.

Please note that the auth-key shown above is the original auth-key. So, the same value of auth-key and Content-Type should be used while configuring Webhook as shown.

  1. Response Mappings: Response Mappings are created via a pairing of a Response Attribute. The Response attribute depends on the output you are returning from the Custom Service JavaScript function. It is recommended to return JSON from the custom service so that you can easily map it with the Marketo field.