Overview of Scoring

Scoring forms the foundation of your collections activities. Oracle Advanced Collections uses scoring in two ways:

Once you know which transactions are delinquent and you know the relative value of a customer, you can manage delinquencies more effectively with either collections strategies or dunning plans.

Note: The score displayed in the Collections Score field in the Collections header is always the customer-level score.

How Customer Value Scoring Works

Let's say you want to evaluate your customers and select the appropriate collections management method for each. In this example, you want to use three factors to select the collections plan for a customer: how much they owe, how many transaction are overdue, and how long they have been a customer. To find out this information, you create a scoring engine with three scoring components to calculate the following information for each customer:

But how do you use the results of these scoring components to rank your customers? For some components, a high number means you are dealing with a good customer, but in other instances, a low number is better.

To qualify the numbers returned by the scoring components, break down the possible numerical results into ranges and assign a value, or score, to each range.

The following tables show the ranges and scores for each scoring component.

Amount Overdue Scoring Component

Amount Overdue Score
$0 - $999 100
$1000 - $4999 50
$5000 - $9999 25
Over $10,000 10

Number of Delinquencies Scoring Component

Number of Delinquencies Score
0 - 9 100
10 - 19 50
20 - 39 25
over 40 10

Customer Since Scoring Component

Customer Since Score
Less than 1 year 10
1 - 2 years 50
3 - 5 years 75
over 5 years 100

However, one type of information about a customer may be more important than another when evaluating which collections activities to use. Let's say that the amount a customer owes is a more important factor than how long they have been your customer or the number of delinquencies they have. You need a way to compare the disparate numbers derived by the scoring components. To do this, assign a weight to each component to indicate its relative importance to the total score. The total weight of all components must add up to 1.0.

The following table shows the weights for the scoring components.

Scoring Component Weight
Amount overdue .5
Number of delinquencies .3
Customer since .2

Based on this setup, Advanced Collections calculates the score for a customer who owes $18,425 on 9 delinquencies, and has been a customer for 2 years as follows:

(10 x .5) + (100 x .3) + (50 x .2) = 5 + 30 + 25 = 60

You can create new scoring engines or use the preconfigured scoring engines provided by Advanced Collections. In either case, you must test your assumptions by running scoring in a test environment with a small segment of your actual data.

How Transaction Scoring Works

Scoring transactions to determine whether they're current, delinquent or pre-delinquent works the same as scoring for customer value, but with one additional step. A transaction scoring engine with one or more scoring components obtains scores for all transactions. Once the transactions have been scored, you need to set score ranges for the status of either current, delinquent or (if your business rules require) pre-delinquent.

For example, your Transaction scoring engine uses a component that looks at the due date of a transaction and then assigns a status based on the score and the defined score ranges. If the due date for a transaction is greater than today's date, it assigns a score of 1 and the transaction is current. If the due date is less than today's date, score is 100 and the transaction is delinquent. Using the preconfigured Delinquency Status Determination scoring engine, the score ranges are shown in the following table:

Score Range Low Score Range High Status
1.00 10.99 Current
11.00 100.00 Delinquent

Related Topics