[Skip Global Navigation]

Customer Success Stories

Success Stories Home

Chelsea Building Society

Using predictive analytics to improve mortgage retention

The situation

Chelsea Building Society is the UK's sixth largest, with assets in excess of £7.9 billion. It has 500,000 saving accounts, over 80,000 borrowers, and a network of 35 branch offices across southern and central England.

The Society has a strong national mortgage sales network to promote its wide‑ranging portfolio of mortgage products across the whole of the UK. The Society employs 943 staff, including 643 based in the Cheltenham Head Office.

As a mutual building society, the Chelsea is owned by its members. Unlike a plc, it doesn't have to pay dividends to shareholders, so it can return more profit to its members by offering better value products, including higher rates for savers and lower mortgage rates for borrowers.

The challenge

Chelsea Building Society has been using software from SPSS Inc. for analytics since 1988. In recent years, the retention of mortgage customers has become an important issue for the Society, and it was recognised that further analytical solutions were needed. Following a presentation Chelsea decided to purchase IBM SPSS Modeler* to improve the borrower retention process.

IBM SPSS Modeler was installed in March and, with the help of consultancy from SPSS Inc., a comprehensive borrower retention model was soon developed for future use.

The results

The results have been very encouraging, and the number of borrowers retained this year has increased by 147 percent. The borrower retention model means that Chelsea Building Society have been able to formulate appropriate contact strategies for those borrowers with a high propensity to redeem – and also those with a high propensity to remain in other accounts. The ensuing cost savings easily outweigh the expenditure on IBM SPSS Modeler.

The solution is viewed internally as a tremendous success, and Chelsea is now looking at extending the use of predictive analytics to other areas such as retail savings. As a result of this success, Chelsea Building Society is now recruiting extra staff to help with the next phase of the project.

Interested in using predictive analytics to improve mortgage retention? Download the Chelsea Building Society PDF here.

*IBM SPSS Modeler, formerly called Clementine®, is part of SPSS Inc.’s Predictive Analytics Software portfolio.

back to top