Scoring and score values

What is scoring?


The word “score” means “to make a hit, win a point or have success”. Score values supply predictions, e.g. on the probability of a consumer defaulting on their payments, for consumers who are not known for dishonouring their payments. This data is determined within the scope of a statistic analysis in order to predict the probability of a customer's defaulting on their payments.

In a scoring procedure, experience from the past is used to predict similar events in the present and future. Similar methods have been used for a long time in market and opinion research, for example in order to forecast election results.

Scoring is an analytical method for forecasting and point evaluation, which calculates the probability with which each individual retail customer will display a certain behaviour, whereby scoring can help to recognize the opportunities and risks of a customer relationship at an early stage. Using this method of analysis, the optimum selection of criteria and weighting is determined in a so-called scorecard, with which companies can best predict the probability of a certain future behaviour by a retail customer.

One of the fields of application is credit scoring, which can be used for all types of lending transactions. Banks use it to assess the credit rating of a borrower, i.e. their ability to fulfil the financial obligations they have undertaken. In the retail market, such analyses and the use of standardized methods are stipulated, in part by the implementation of the Basel II guidelines.

The criteria, i.e. the data used in scorecards such as the Informa-Consumer-Score or industrial scores and their weighting are strongly dependent on the industry applying them (e.g. the mail-order business, credit services sector, telecommunications incl. mobile telephone systems, insurance) and the objective of the prediction (e.g. the risk of default in payment or the inclination to cancel).

In some sectors, only little valid data is available. In such cases, external data is often used for making a decision, for example from marketing data, risk data and external score value providers. However, credit-relevant criteria are always used, the predictive quality of which has been validated for each inquirer in comprehensive mathematical analyses.

For example, no data is included on the external assessment of houses or addresses, but instead, for example, entries in public directories and registers such as the debtors' register or the register of companies, in an aggregated form. This information is processed on the basis of data protection regulations.

Why implement scoring?

When based on scorecards, companies reach decisions that are more objective and efficient, work more cost-effectively and can, therefore, offer their customers more competitive prices and better terms. Especially in volume business such as the mail-order business, this simplifies and speeds up economic decision-making processes for both the companies and their customers. It enables the greatest number of transactions to be concluded with the least possible risk while, at the same time, preventing all customers from having to pay increased costs due to the default in payment of individual customers.