A scoring model is one of the primary credit review tools that Oracle Credit Management uses to assess the creditworthiness of your customers and prospects. See: Overview of Oracle Credit Management.
Scoring models include the data points and scoring method that are appropriate for a particular credit review. When defining scoring models, for each data point, (1) indicate a score for each range of values and (2) optionally assign a relative weighting factor.
During a credit analysis, Credit Management uses the scores that you assigned to each data point range of values to calculate a score. The lower the credit score, the greater the credit risk.
Note: Credit scores are calculated as whole integers, with decimal values .5 or greater rounded up. For example, 70.5 becomes 71, and 70.4 becomes 70.
The power of the scoring model lies in its ability to facilitate the automated credit review process, because you can attach recommendations to each score range. Once a score has been calculated, Credit Management can automatically determine the appropriate credit recommendations without user intervention. See: Assigning Automation Rules.
Credit Management provides you with the ability to define multiple scoring models. Define a different scoring model for each type of credit review that you need. This lets you apply standard and consistent scoring guidelines across your customers and prospects.
You can assign a default scoring model to a credit checklist. See: Defining Checklists.
Later, during credit analysis, you can generate various "what-if" scenarios during a credit review by selecting different scoring models from the case folder. See: Calculating a Credit Score.
Suggestion: You can ignore Credit Management's scoring model altogether and use an external scoring engine to derive a credit score. See: External Scoring.
You can assign different scoring attributes to each data point in your scoring model.
For each data point that the scoring model includes, you must assign a range of values and a corresponding score for each range. The ranges of values for a data point typically represent levels of credit risk.
Numeric ranges can be positive or negative, and you can use decimal points. Scores must be numeric values, either positive or negative.
For example, this table illustrates sample ranges and scores for the Percentage of Invoices Paid Late data point:
| Credit Risk | Range Value | Score |
|---|---|---|
| Low | 0 to 20 | 15 |
| Moderate | 21 to 50 | 10 |
| High | 51 to 100 | 0 |
This table illustrates sample ranges and scores for the DSO data point:
| Credit Risk | Range Value | Score |
|---|---|---|
| Low | -999 to 9 | 8 |
| Moderate | 10 to 24 | 5 |
| High | 25 to 34 | 2 |
| Highest | 35 to 999 | 1 |
Assign a weighting factor to each data point to indicate the relative importance of each data point in the scoring model.
For example, will you use this scoring model to determine a credit limit increase for an existing customer who has years of credit history with your enterprise? In that case, you might assign a higher weighting factor to the Percentages Of Invoices Paid Late data point, and a lower weighting factor to the Credit Agency Score data point.
Or, assign no weights at all to produce a raw score. If no weights are assigned, then the calculated score could fall outside the 0 to 100 range.
Suggestion: You can view the minimum and maximum potential raw scores on the Review page during scoring model definition. This is helpful to know when defining automation rules. See: Assigning Automation Rules.
Continuing the previous example, this table illustrates the possible weighting factors for the Percentage of Invoices Paid Late and DSO data points:
| Data Point | Weight |
|---|---|
| Percentage of Invoices Paid Late | 65 |
| DSO | 35 |
Note: If you assign weights, then the sum of all data point weighting factors must equal 100.
This example illustrates how the credit score is calculated, based on the scoring model defined in the previous examples. This table shows the values for each data point.
| Data Point | Value |
|---|---|
| Percentage of Invoices Paid Late | 75 |
| DSO | 15 |
For each data point, the score for the value's range is divided by the largest possible score for the data point.
Percentage of Invoices Paid Late: 0 / 15 = 0
DSO: 5 / 8 = .625
The result from step 1 is multiplied by the weight.
Percentage of Invoices Paid Late: 0 * .65 = 0
DSO: .625 * .35 = .13671875
The results of step 1 and 2 are added for each data point.
0 + .13671875 = .13671875
The result of step 3 is multiplied by 100.
.13671875 * 100 = 13.671875
The credit score is 14.
On the Define Scoring Model page, enter the name and description of this scoring model.
In the Currency field, from the list of values, select the currency for this scoring model.
The selected currency indicates the currency in which the data point range is expressed. Credit Management uses the Exchange Rate Type system option during currency conversion.
Use the Notes field to optionally enter comments about this scoring model.
The Start Date defaults to the current date, but you can change it to a future date.
If your credit requirements change and you want to define a new scoring model that is more in line with your revised credit policies, inactivate the outdated scoring model by entering an end date. The only valid end date is the current date.
Suggestion: Inactive scoring models do not display on the scoring models search page, unless you select the Show Inactive Scoring Model check box as a search parameter.
When you create a new case folder or attach a scoring model to a credit checklist, you can select only a scoring model that has no end date, or an end date that is greater than the case folder or checklist creation date.
Note: Once you enter an end date and save your work, you can no longer change or remove the date. This ensures that your credit policies are strictly enforced, and also that comparisons across scoring models remain meaningful.
Use the Convert null values to zero values check box to indicate how Credit Management should treat data points with null values.
Suggestion: Carefully consider your business requirements before leaving this box blank. If you do not select this box, then a scoring model will not calculate a score if a data point contains a null value. In this case, the workflow will stop and the credit review will be routed to the Credit Scheduler for credit analyst assignment.
On the Select Data Points page, select the data points that you want to include in the scoring model.
For each data point that is included in this scoring model, assign a range of values and a corresponding score for each range.
The ranges that you enter should include all possible values. You can enter up to 15 characters for each value in the range.
Note: If you enter alphanumeric values for a data point, then both range values must be the same. This table illustrates an example:
| Range | Low | High | Score |
|---|---|---|---|
| Range 1 | A2B | A2B | 10 |
| Range 2 | B3B | B3B | 5 |
When you have completed assigning ranges and scores to the data points in this scoring model, optionally assign a weight to each data point. The weight that you assign indicates the relative importance of the data point in the scoring model.
Review the scoring model.
Suggestion: If you have not assigned weights, then the scoring model will calculate a raw score. Note the minimum and maximum potential raw scores on this page, because you will need to know this range if you define automation rules. See: Assigning Automation Rules.
You can still update a scoring model after you save it. After you submit a scoring model, however, you can no longer update it.
What is the difference between a null data point value and a zero data point value?
A null data point value indicates no activity, while a zero data point value indicates that activity has occurred in the past, but there is no value for the particular data point. For example, let's consider the Count of Invoices Paid Late data point. In this case, a null value might indicate that no payments have been made, while a zero value indicates that all invoices were paid on time.