Predictive model

Predictive modeling is a statistical technique used to predict the likelihood of an outcome.

For example, the success of a sale may depend on certain factors, such as the value of a sale or a product type. Using a predictive model, you can analyze data from within a process and use the insight to prioritize tasks and make decisions to achieve the optimal outcome for your business.

For example, in an insurance company, a process is dedicated to selling insurance. From previous experience, the company knows the success rate is higher when younger males are targeted. To increase the chances of meeting the sales target before the end of the quarter, the company defines a model to calculate a score that helps prioritize where the insurance can be sold successfully.

Each model consists of several variables (predictors) that are likely to influence the outcome, and the corresponding weighting and scoring to work out the statistical model. See Manage a predictive model.

Define three process variables: Score (Long), Gender (String) and Age (Long).
Create a model based on Gender and Age, to calculate score when the age falls between 25 and 30 years.

Define value/range and weight for the variables as given in the following table.

You can add variables multiple times and give a different weight depending on the value or range.

Process variable	Value/Range	Weight
Gender	Male	2
Gender	Female	1
Age	Less Than 25	4
Age	Greater Than 25 and Less Than or Equal To 30	3
Age	Greater Than 30 and Less Than or Equal To 50	1

Define the scoring system by creating a score rule as: Gender + (Age * 2)
Thus, the score for a male, aged 26 is 8 and the score for a male, aged 18 is 10. This resulting score can be used in decision logic and work allocation rules to direct high-scoring insurance requests to the right reviewers, thus increasing potential revenue to the company.