The Deutsche Telekom Group is one of the world’s leading companies in the telecommunication sector and has 106.4 million customers worldwide as part of its T-Mobile operations, 33.2million of which are in Germany. With 10 million broadband customers and 33.2 million fixed line customers in Germany, the overall revenue of the group as a whole reached €61.3 billion in 2006.

SAF is the business network of T-Systems and the business customer branch of Deutsche Telekom. It provides a range of risk management, collections management and credit bureau services to many important customers in Germany, including the T-Group, a large German energy company, plus several national financial institutions. With ten million credit assessments, two million assigned collection cases and two million processed address identifications per year, SAF belongs to the top three of its industry in Germany.

The services offered by SAF include address verification, credit check (B2C & B2B), scoring and collections services. The collections industry, in particular, is highly profitable and is still growing, although the German collections market is strongly fragmented and is predominantly automated and IT-controlled. However, due to high margins and levels of potential growth, the collections market is an attractive revenue-generation prospect. The collections process itself fits in with the SAF delivery structure as part of its Receivable Management division, presently generating 73% of SAF’s revenue.

The situation in Germany is that the number of insolvencies has been growing at a considerable rate for most of the last decade. This has resulted in T-System’s view that in order to find the right customer you have to select them carefully. Through scoring, credit checking and customer detail verification, SAF is able to select those ‘right’ customers, differentiating between those that are expected to pay, have the potential to default with their payments, or are simply a ‘non-customer’, for example those identified as criminals.

The traditional approach to scoring uses negative information attached to a person’s identity. When the database is searched, individuals with known negative behaviour can be identified. Typically, the share of people with negative creditworthiness is about 5-10% but there is no further differentiation of the remaining 90-95% with respect to credit risk. The challenge here is how can we reduce the 90-95% volume, and how can potential risks of people without negative features be assessed?

Unfortunately, no German credit agency has data sources which cover the entire process of the ‘debtor career’. ‘Hard’ negative customers (with court judgments or on insolvency lists) are identified mostly by all credit agencies, but there are remarkable differences in relation to the consistency of this data. Not all credit agencies have an automatic link to district courts and/or the federal legal gazette. ‘Soft’ and ‘medium’ negative customers (consisting of extra-judicial and/or judicial collection procedures) are not identified by all credit agencies. As a result, many inquiry agencies cannot answer the inquiries with 100% accuracy.

Our solution to this problem is SAFE - a joint project between SAF and Experian Decision Analytics. We are able to draw upon the data of 40 million fixed line telecoms customers, 15 million mobile telecoms customers, 70 million customers of telecommunications, energy supply, banks and mail order, plus an additional 60 million in the near future as a result of integration of a co-operation partner’s database. This will give us a wealth of personal data, individual payment behaviour, negative data and micro-geographic data, the use of which is in line with current data protection laws.

The SAFE consumer score was developed from the SAF Group database, specifically the debtors register, collections cases, master file data, payment records and the insolvency database. The SAF Group database is unique with regards to the coverage of, and the information it holds about, German households. It has almost 40 million single data records, each with 100 criteria, and has 95% coverage of all German streets.

The SAFE Consumer Score closes the information gap between people with negative creditworthiness and people without negative data. It quantifies the credit risk of people for whom no negative data is held. All credit inquiries are answered completely and, as a result, the SAFE Consumer Score presents the probability of payment default of a customer based on this information.

SAF is able to use the scoring suite for marketing purposes, rating the risk of the new customer segment, to support the internal score in the existing customer segment, and as part of the collections procedure. Experian Decision Analytics conducted a credit bureau benchmark to identify the optimal mix of credit bureau information for a client to improve its New Business Management. A data sample of historic applications with a known customer profile was matched with credit bureau data retrospectively at application date and today (date of sampling). The objective of this was to evaluate existing credit bureau information in order to identify defaulters. The results were that the best combination of external data sources to predict defaulted applicants was identified, and the incremental use of credit bureau scores in a bespoke application scorecard has led to quality improvement of 20% (Gini coefficient).

In conclusion, the risk costs could be improved by 17% using an optimal set of external data and an improved scoring model.

Thomas Grothe – CBDO, SAF Forderungsmanagement GmbH & Holger Dittombee – Experian Decision Analytics - From their presentation given at the Experian Decision Analytics European Telco Forum 2007

Contact us for further discussions about this article

View this article in German

Top