Scoring

Precise risk management

Request a quote
  1. Home
     > 
  2. Solutions
     > 
  3. Risk Management
     > 
  4. Scoring

Scoring - Our Scheme

Know in Advance What Your Customers Will Do

Predict the Behavior of Your Customers

The ascertainment of a score value optimally complements acredit check. To achieve this, statistical methods are employed to calculate the probability of payment based on data such as age, place of residence, and time of order placement. The obtained results are integrated into the detailed recommendations that we provide you.

Essential Points about Scoring

Scoring is an analytical procedure for making predictions, which calculates the probability with which your customer will fulfill payment. The result is a so-called scorecard. This represents the set of rules according to which data is weighted and evaluated. Within a category such as age, the system allocates a certain number of points. These points are added together according to their significance and render the respective score value. The categories used are manifold: Micro-geographic data and socio-demographic data can be evaluated as well as the time of placing an order, value of a shopping cart or type of the products. Thus it is possible to predict payment behavior very accurately based on data which the customer gives at the time of purchase.

Ask for an offer now

Creating Professional Scorecards

Individual scorecards tailored to your business and your customers are always superior to standardized scoring applications. Quality is always crucial: if scorecards are not defined sufficiently clearly, subsequent scoring results will be inaccurate. This can lead to a loss of revenue, if too many customers are declined or can drastically increase the default rate. The high quality of our scoring is based on the creation of scorecards, which are used by numerous well-known companies in the e-commerce and mail order business.

To create appropriate scorecards, your customers are analyzed in detail using all possible criteria relevant for making a score-related decision. Thus “historical” knowledge is used to adequately predict future decisions of your customers.

After conducting an explorative data analysis and determining the final target variables (payment behavior), we draw suitable samples for analysis, validation, and testing to develop respective scorecards. All potential scoring characteristics are examined with regard to their individual ability to enable differentiation. An optimal grouping ensures a maximum of information and at the same time stabilizes the forecast capacity of individual scoring characteristics. Using a multivariate non-linear regression model, we then develop the actual scorecard. Scoring characteristics are selected according to an iterative procedure and as well as taking their respective intercorrelations into consideration. To avoid a distortion of scorecards and to improve the forecasting model, we use various analysis models to scrutinize the allocation algorithms for rejected cases. The implemented scorecards are then regularly analyzed by our risk management specialists. They check whether the required differentiation is still warranted or whether an adjustment is necessary.

Ask for an offer now

Solutions for Your New and Existing Customers

The score rating for new customers is called application scoring, and can only be performed utilizing criteria existing at the time of the first order placement. The dynamic enrichment with credit-related information from credit reporting agencies additionally plays an important role, and enables astonishingly accurate forecasts with regard to payment behavior.

An exact scoring scale enables to monitor the exact cut-off value, i.e. the point at which the customer can be rejected.

For existing customers, the scope of criteria relevant for making decisions is much larger, because empirical information about the payment behavior already exists. These criteria are used on hand of behavior scores to predict future behavior. This way you are able to identify “critical customers” early on and align your limit-policy accordingly.

Ask for an offer now

Share now: